East Texas Council of Governments · Shagun Bhat Amnon Bar-Ilan Sue Kemball-Cook Greg Yarwood...

EMISSIONS FROM

NATURAL GAS EXPLORATION AND PRODUCTION ACTIVITY IN THE HAYNESVILLE SHALE

Prepared for: East Texas Council of Governments

3800 Stone Road Kilgore, TX 75662

Prepared by:

Rajashi Parikh John Grant

Shagun Bhat Amnon Bar-Ilan

Sue Kemball-Cook Greg Yarwood

ENVIRON International Corporation 773 San Marin Drive, Suite 2115

Novato, California, 94998 www.environcorp.com

P-415-899-0700 F-415-899-0707

January 2013

06-26410C1

January 2013

i

CONTENTS

EXECUTIVE SUMMARY ......................................................................................................... 6

1.0 INTRODUCTION ........................................................................................................... 10

1.1 Shale Activity in the U.S. ................................................................................................ 12

1.2 The Haynesville Shale .................................................................................................... 13

1.2.1 Previous Haynesville Emission Inventory Study ................................................. 13

1.2.2 Haynesville Development Since the Previous Haynesville Shale Inventory ............................................................................................................. 16

2.0 OBJECTIVE OF THIS STUDY ........................................................................................... 18

3.0 SCOPE OF STUDY .......................................................................................................... 19

3.1 Pollutants ....................................................................................................................... 19

3.2 Geographic Extent of the Haynesville Shale.................................................................. 19

3.3 Temporal Scope ............................................................................................................. 24

4.0 METHODOLOGY ........................................................................................................... 25

4.1 Survey ............................................................................................................................ 25

4.1.1 Survey Outreach Process and Results ................................................................. 25

4.2 PRODUCTION STATISTICS DEVELOPMENT .................................................................... 28

4.2.1 Drilling (Spuds) Estimates ................................................................................... 28

4.2.2 Well Counts and Production Estimates .............................................................. 29

4.3 Permitted Point Sources ................................................................................................ 30

4.4 Methodology for Base Year Emissions .......................................................................... 32

4.4.1 Well Pad Construction ........................................................................................ 33

4.4.2 Drill Rigs – Drilling Operations ............................................................................ 35

4.4.3 Fracing ................................................................................................................. 36

4.4.4 Completion Venting ............................................................................................ 37

4.4.5 Blowdown Venting .............................................................................................. 39

4.4.6 Fugitives (Leaks) .................................................................................................. 40

4.4.7 Pneumatic Devices .............................................................................................. 42

4.4.8 Heaters ................................................................................................................ 44

4.4.9 Dehydrators ........................................................................................................ 45

4.4.10 Flaring 46

4.4.11 Wellhead Compressors ....................................................................................... 47

January 2013

ii

4.4.12 Condensate Tanks ............................................................................................... 49

4.4.13 Midstream/Gas Processing Sources ................................................................... 50

4.5 Forecasting Scenarios .................................................................................................... 51

4.5.1 Rig Count Estimates ............................................................................................ 52

4.5.2 Drilling and Well Count Estimates ...................................................................... 55

4.5.3 Production Estimates .......................................................................................... 57

4.6 Future Year Control Methodology ................................................................................ 62

4.6.1 Nonroad Diesel Engine Standards and Fuel Sulfur Standards ............................ 63

4.6.2 Spark Ignition Engines ......................................................................................... 63

4.6.3 Subparts OOOO: New Source Performance Standards regulating VOC, SO2 and new affected facilities ........................................................................... 65

5.0 RESULTS....................................................................................................................... 69

5.1 Detailed Emissions Inventory for 2012 ......................................................................... 69

5.2 Emissions Inventories for Calendar Years 2011-2020 ................................................... 74

5.3 Comparison of Emissions............................................................................................... 79

5.4 Discussion ...................................................................................................................... 82

6.0 NEXT STEPS .................................................................................................................. 83

7.0 REFERENCES ................................................................................................................ 84

APPENDICES Appendix A. Letter and Questionnaire Sent to Haynesville Shale Producers

TABLES

Table ES-1. 2011 to 2020 emissions of NOx, VOC, and CO by scenario for the Haynesville Shale region.......................................................................................... 9

Table 1-1. U.S. shale gas production 2007-2010. ................................................................... 12

Table 3-1. Counties in Texas and parishes in Louisiana considered part of the Haynesville Shale development area. ................................................................... 21

Table 3-2. Counties in Texas and parishes in Louisiana considered part of the Haynesville Shale midstream gas gathering and processing area. ....................... 23

Table 4-1. 2011 Haynesville Shale gas production ownership by surveyed producers. ............................................................................................................. 26

Table 4-2. Comparison of gas composition analyses used in original and updated emission inventories. ............................................................................................ 28

January 2013

iii

Table 4-3. Summary of Northeast Texas and Northwest Louisiana Active Well Counts for Years 2008-2011. ................................................................................. 30

Table 4-4. Summary of Base Year Gas and Condensate Production for the Haynesville Region. ............................................................................................... 30

Table 4-5. 2010 gas production in all Texas counties and Louisiana parishes considered part of the Haynesville Shale midstream/gas processing region..................................................................................................................... 31

Table 4-6. Haynesville Shale emission inventory source categories. ..................................... 32

Table 4-7. By source category scaling parameter. ................................................................. 33

Table 4-8. Well pad construction emissions estimation inputs. ............................................ 34

Table 4-9. Drill rig emissions estimation inputs. .................................................................... 36

Table 4-10. Fracing emissions estimation inputs. .................................................................... 37

Table 4-11. Completion venting emissions estimation inputs. ................................................ 39

Table 4-12. Blowdown venting emissions estimation inputs. ................................................. 40

Table 4-13. Fugitives emissions estimation inputs. ................................................................. 42

Table 4-14. Pneumatic devices emissions estimation inputs. ................................................. 44

Table 4-15. Heater emissions estimation inputs. ..................................................................... 45

Table 4-16. Dehydrator emissions estimation inputs. ............................................................. 46

Table 4-17. Flaring emissions estimation inputs. ..................................................................... 47

Table 4-18. Wellhead compressor emissions estimation inputs. ............................................ 49

Table 4-19. Condensate tanks emissions input. ....................................................................... 50

Table 4-20. Low scenario Haynesville Shale rig count estimates. .............................................. 53

Table 4-21. Summary of federal and state “on-the-books” regulations affecting the oil and gas source categories considered in this inventory. ................................. 63

Table 4-22. Federal NSPS emissions standards for engines less than 25 horsepower. .......................................................................................................... 64

Table 4-23. Federal NSPS emissions standards for engines greater than 25 horsepower but less than 100 horsepower. ......................................................... 64

Table 4-24. Federal NSPS emissions standards for engines greater than 25 horsepower but less than 100 horsepower. ......................................................... 64

Table 4-25. Subpart OOOO requirements for sources at the well site. ................................... 67

Table 4-26. Subpart OOOO requirements for sources at natural gas processing plants. .................................................................................................................... 68

January 2013

iv

Table 5-1. 2012 emissions (TPD) of NOx, VOC, and CO by scenario for the Haynesville Shale formation. ................................................................................. 70

Table 5-2. 2011 to 2020 emissions of NOx, VOC, and CO by scenario for the Haynesville Shale region and percent difference from the moderate scenario. ................................................................................................................ 75

Table 5-3. Percentage change in per well base year emissions from previous ENVIRON (2009) and the updated emission inventory. ........................................ 79

Table 5-4. Original ENVIRON (2009) and updated emission inventory presented in this report for all scenarios (tons per day). ........................................................... 80

Table 5-5. Summary of Haynesville base year emissions and 2009 Barnett Shale emissions with regional production statistics. ...................................................... 81

Table 5-6. By source category emissions for Barnett Shale and Haynesville Shale emission inventory. ............................................................................................... 82

FIGURES

Figure ES-1. 2012 moderate development scenario Haynesville Shale formation NOx proportional contributions by source category. ............................................. 8

Figure ES-2. 2012 moderate development scenario Haynesville Shale formation VOC proportional contributions by source category. ............................................. 8

Figure 1-1. Schematic of Haynesville Shale well showing horizontal drilling and hydraulic fracturing techniques. Image from http://geology.com/articles/haynesville-shale.shtml. ......................................... 10

Figure 1-2. Haynesville and Barnett Shale gas production (bcf/day) (EIA, 2011b). ................ 11

Figure 1-3. Major U.S. shale plays in the lower 48 states (EIA, 2011c). .................................. 12

Figure 1-4. U.S Shale Gas Production (bcf/day) from 2000 to 2012 (EIA, 2012)..................... 13

Figure 1-5. Count of drill rigs active in the Haynesville Shale from January 2010-December 2012. Figure from http://www.haynesvilleplay.com/. ...................... 16

Figure 1-6. Projected annual average drill rig count for low, medium and high scenarios from ENVIRON (2009). Blue bars show actual drill rig counts for March of each year from 2009-2012. Drill rig counts from http://www.haynesvilleplay.com/. ....................................................................... 17

Figure 3-1. Upper left panel: Spatial extent of the Haynesville Shale in Texas by county and well locations as of February, 2009. Upper right panel: Spatial extent of the Haynesville Shale in Texas by county and well locations as of December, 2012. Figures from the Railroad Commission of Texas website (TRRC, 2012). Pale blue shading indicates that the TRRC considers that county to be a core Haynesville County. Yellow shading indicates that the TRRC considers that county

January 2013

v

to be a non-core Haynesville County. Lower left panel: Extent of the Haynesville Shale in Louisiana as of February, 2009 (LDNR, 2012c). Lower right panel: Extent of the Haynesville Shale in Louisiana as of December, 2012 (LDNR, 2012c). ........................................................................... 20

Figure 3-2. Spatial extent of the Haynesville Shale in Texas and Louisiana as defined in this study. ............................................................................................. 22

Figure 3-3. Midstream facilities in Texas and Louisiana (identified from state emission inventory databases) in the Haynesville Shale and surrounding region considered in this analysis. .................................................... 24

Figure 4-1. Forecast of future rig count based on 2012 rig count trends. .............................. 53

Figure 4-2. Total annual drill rig count in the Haynesville Shale predicted for the three development scenarios in the period 2009-2020. ...................................... 55

Figure 4-3. Haynesville shale annual spud counts by year for each scenario. ........................ 56

Figure 4-4. Haynesville shale active well count estimates by year for each scenario. ................................................................................................................ 56

Figure 4-5. Monthly gas production for well 236242 (LDNR, 2012d). .................................... 57

Figure 4-6. Average percentage decline from the first half year of gas production. .............. 58

Figure 4-7. Haynesville Shale well decline estimate. .............................................................. 59

Figure 4-8. Haynesville Shale well decline estimate with confidence intervals. ..................... 60

Figure 4-9. Total Haynesville Shale annual gas production estimates. ................................... 61

Figure 4-10. Total Haynesville Shale cumulative gas production estimates. ............................ 61

Figure 4-11. Total Haynesville Shale annual condensate production estimates. ..................... 62

Figure 5-1. 2012 moderate development scenario Haynesville Shale formation NOx proportional contributions by source category. ........................................... 71

Figure 5-2. 2012 moderate development scenario Haynesville Shale formation VOC proportional contributions by source category. ........................................... 72

Figure 5-3. 2012 moderate scenario Haynesville Shale formation CO proportional contributions by source category. ........................................................................ 73

Figure 5-4. 2012 Haynesville Shale formation emissions of NOx, VOC, and CO by scenario and source category. .............................................................................. 74

Figure 5-5. 2011 to 2020 moderate scenario Haynesville Shale NOx emissions by source category. .................................................................................................... 76

Figure 5-6. 2011 to 2020 moderate scenario Haynesville Shale VOC emissions by source category. .................................................................................................... 77

Figure 5-7. 2011 to 2020 moderate scenario Haynesville Shale CO emissions by source category. .................................................................................................... 78

January 2013

6

EXECUTIVE SUMMARY

The Haynesville Shale is located approximately 10,000-13,000 feet beneath Northeast Texas and Northwest Louisiana and contains very large recoverable reserves of natural gas (EIA, 2011). Intensive exploration of the Haynesville began in 2008, and as of December 2012, there are nearly 3,000 wells producing natural gas from this formation. The development of natural gas resources in the Haynesville has been economically important, but may also generate significant emissions of ozone precursors in a region with several ozone monitors that are close to or exceed the 2008 National Ambient Air Quality Standard.

During 2009, Northeast Texas Air Care (NETAC; www.netac.org) developed an emission inventory of ozone precursors for projected future Haynesville Shale development from 2009 through 2020. Using well production data from state regulatory agencies and a review of the available literature, projections of future year Haynesville Shale natural gas production were derived for 2009-2020 for three scenarios corresponding to limited, moderate, and aggressive development (ENVIRON, 2009). These production estimates were then used to develop an emission inventory for each of the three scenarios. The emission inventory covered 5 Texas counties and 6 Louisiana parishes. NETAC’s photochemical modeling of the year 2012 using this emission inventory showed that 8-hour ozone impacts occurred within Northeast Texas and Northwest Louisiana as a result of development in the Haynesville Shale, with projected ozone design value increases ranging from 1-5 ppb at Northeast Texas ozone monitors for the aggressive development scenario. Modeled ozone increases due to Haynesville Shale emissions also affected regions outside Northeast Texas and Northwest Louisiana due to ozone transport.

2009 and 2010 were years of rapid development in the Haynesville. In March 2010, there were over 160 rigs drilling in the Haynesville; this far exceeded even the aggressive scenario prediction of the ENVIRON (2009) study. Since 2010, the pace of development has slowed due to low natural gas prices and high oil-to-gas price ratio, which encouraged development of formations that contain hydrocarbon liquids, such as the Eagle Ford Shale. As of December 2012, there are approximately 24 rigs active in the Haynesville; this figure is far below the ENVIRON (2009) low scenario prediction of 95 rigs.

Given the large number of active, producing wells, the Haynesville Shale continues to be an emissions source that must be evaluated and accurately represented in NETAC’s ozone modeling. The divergence between the predicted and actual drill rig activity and well count as well as the expanded spatial extent of the Haynesville Shale (now covering 10 Texas counties and 8 Louisiana parishes) indicate that the future year projections made in the ENVIRON (2009) study are now out of date, and that an updated emission inventory is necessary. This report describes the development of a revised Haynesville Shale inventory that takes advantage of a longer well production data record as well as survey data from Haynesville Shale producers.

Projections of drilling activity and gas production in the Haynesville Shale were developed for each year for the period 2012-2020 using well production data from state regulatory agencies and a review of the available literature. These projections form the basis for scaling the 2011 base year emission inventory to all future years. The projections were developed for three scenarios: (1) a moderate (medium) growth scenario; (2) a low growth scenario; and (3) an

January 2013

7

aggressive (high) development scenario. These scenarios were developed to cover a range of possible future levels of gas exploration and production activity in the Haynesville Shale.

The projections were then used to develop a 2011 base year inventory and future year annual emissions inventories for ozone precursors for each year from 2012-2020 for all three scenarios. A detailed emission inventory for photochemical modeling was developed for all three scenarios for the year 2012, so that a revised ozone impact assessment can be performed with the updated Haynesville Shale emission inventory. In the updated emission inventory, 2012 nitrogen oxide (NOx) emissions from the development of the Haynesville Shale formation ranged from 32 tons/day for the low and moderate scenarios to 50 tons/day for the high scenario. 2012 NOx and volatile organic compound (VOC) emissions are shown in Figure ES-1 and Figure ES-2 respectively. Estimates of 2020 NOx emissions ranged from 25 tons/day in the low development scenario to 46 tons/day in the moderate scenario to 149 tons/day in the high scenario (Table ES-1). The magnitude of these NOx emissions estimates indicates that the Haynesville Shale continues to be a significant source of emissions that can affect ozone in Northeast Texas.

Revisions to key emission inventory data including well drilling time, gas composition analysis, wellhead compressor engine prevalence, and well decline and projection scenarios have resulted in an updated Haynesville Shale emission inventory that differs significantly from the original ENVIRON (2009) inventory. NOx and CO emissions for all scenarios are lower in the updated emission inventory relative to the original emission inventory. VOC emissions for all scenarios are higher in the updated emission inventory relative to the original emission inventory. In the updated inventory, there is a reduction in emissions from drilling, well site heaters and wellhead compressor engines. The activity data used for venting sources were unchanged from the original emission inventory except for dehydrators and fugitives, for which assumptions were updated. VOC emissions from most of the venting source categories were reduced due to the lower average VOC content of the vented gas based on the Haynesville specific gas composition analysis relative to the VOC content of the gas composition used in the original ENVIRON (2009) study. Higher VOC emissions from dehydrators and fugitives relative to the original emission inventory (ENVIRON, 2009) are due to updated assumptions for these two source categories. The well decline curve estimates of gas production are approximately three times higher over the first 20 years of production while the point source emissions per unit gas production were reduced by 79%, 64%, and 81% for NOx, VOC and CO, respectively relative to original ENVIRON (2009) study.

There is significant uncertainty associated with the future year emissions estimates since development of the Haynesville Shale will depend on economic and environmental factors that are difficult to forecast. The low and high scenarios are meant to represent the likely range of emissions from the development of the Haynesville Shale formation. Analysis of the emission inventories suggests that if the development of the Haynesville Shale continues at even a relatively slow pace, emissions from exploration and production activities will be sufficiently large that their potential impacts on ozone levels in Northeast Texas should continue to be evaluated.

January 2013

8

Figure ES-1. 2012 moderate development scenario Haynesville Shale formation NOx proportional contributions by source category.

Figure ES-2. 2012 moderate development scenario Haynesville Shale formation VOC proportional contributions by source category.

2012 Dehydrator0.65%

2012 Drill Rigs15.99%

2012 Flaring0.01%

2012 Fracing1.89%

2012 Heaters1.23%

2012 Midstream CS and NGP

79.70%

2012 Wellhead Compressors

0.02%

2012 Wellpad Construction

0.51%

2012 Haynesville Shale NOx Emissions Contribution By Source Category for Moderate Scenario

Midstream CS and NGP are Midstream Compressor Stations and Natural Gas Plants

2012 Blowdowns

0.31%

2012 Completion Venting6.72%

2012 Dehydrator42.60% 2012 Drill Rigs

1.17%

2012 Fracing0.39%

2012 Fugitives0.37%

2012 Heaters0.09%


40.41%

2012 Pneumatic

Devices6.22%


0.02%


0.09%

2012 Condensate

Tank1.61%

2012 Haynseville Shale VOC Emissions Contribution By Source Category for Moderate Scenario


January 2013

9

Table ES-1. 2011 to 2020 emissions of NOx, VOC, and CO by scenario for the Haynesville Shale region.

Year

NOx Emissions (TPD) VOC Emissions (TPD) CO Emissions (TPD)

Low Moderate High Low Moderate High Low Moderate High

2011 35.5 35.5 35.5 23.3 23.3 23.3 21.8 21.8 21.8

2012 31.5 31.5 50.1 24.3 24.3 35.7 19.6 19.6 30.8

2013 28.6 28.6 70.0 23.4 23.4 52.8 17.9 17.9 43.1

2014 26.4 26.4 86.2 22.0 22.0 66.9 16.5 16.5 53.4

2015 25.4 25.4 99.4 21.5 21.5 78.7 15.9 15.9 61.7

2016 24.9 24.9 110.9 21.3 21.3 89.1 15.7 15.7 68.8

2017 24.7 27.3 121.6 21.3 23.2 98.8 15.5 17.1 75.6

2018 24.8 32.5 132.1 21.5 27.4 108.2 15.6 20.2 82.0

2019 24.8 38.9 141.2 21.6 32.7 116.4 15.6 24.1 87.7

2020 24.8 46.2 148.7 21.7 38.7 123.1 15.6 28.4 92.2

January 2013

10

1.0 INTRODUCTION

The Haynesville Shale is an Upper Jurassic-era layer of sedimentary rock that was discovered 60 years ago near the town of Haynesville in Claiborne Parish, Louisiana. The Haynesville Shale is approximately 300 feet thick and lies at depths of 10,000 to 13,000 feet below the surface in Northeast Texas and Northwest Louisiana. The Haynesville Shale is sometimes referred to as the Bossier Shale in Texas. The region was once covered by water and the shale formed as sediments were deposited on the sea floor and were compacted as subsequent layers formed overhead. The natural gas present within the rock is the result of the deposition of organic material along with the sediment and the chemical transformation of this material from the resulting heat and pressure of the overlying layers.

Shale rock has very low permeability, making it difficult for gas to be produced from the rock. Until recent technological advances, it was considered economically infeasible to develop natural gas reserves in shale. Horizontal drilling techniques and advances in hydraulic rock fracturing (fracing) methods have made it profitable to extract natural gas from shale layers.

A horizontal well is shown in Figure 1-1. Initially, the well is drilled vertically, but once the formation of interest has been reached, the drill bit is angled so the well turns and runs horizontally along the formation. This increases the well’s exposure to the formation, enhancing production. Next, the rock surrounding the well bore is fractured hydraulically in order to stimulate production. Water and proppants such as sand or ceramic beads are pumped into the well bore under high pressure and create cracks in the rock through which natural gas can flow into the well bore. The purpose of the proppant is to keep the cracks open so that gas continues to flow to the well.

Figure 1-1. Schematic of Haynesville Shale well showing horizontal drilling and hydraulic fracturing techniques. Image from http://geology.com/articles/haynesville-shale.shtml.

January 2013

11

The Barnett Shale in the Dallas-Fort Worth area of Texas is a shale play that has been an area of intensive natural gas development over the last decade, with large increases in production coming due to widespread adoption of horizontal drilling and fracing methods. The Haynesville Shale is similar to the Barnett Shale in terms of required gas extraction techniques, but the Haynesville is potentially a much larger resource (Groundwater Resources Council et al., 2009). Exploratory wells drilled in the Haynesville Shale area in 2007-8 were determined to be highly productive, and the Haynesville Shale was promoted as a significant new energy resource. For example, a 2008 report in the Wall Street Journal indicated that as many as 15,000 new wells were planned for the Haynesville Shale area (Wall Street Journal, 2008). Development began in the 2008-2009 time period and grew rapidly through 2011 (TRRC, 2012a and LDNR, 2012c). Although recently a slowdown in drilling activity has occurred, the Haynesville Shale is now considered a major gas production area. According to some estimates (Groundwater Resources Council et al., 2009), the Haynesville Shale formation may contain as much as 250 trillion cubic feet (TCF) of recoverable gas, and several major energy companies are developing this formation.

As of December 2012, there are 2,141 active producing Haynesville Shale wells in Louisiana with another 176 wells awaiting completion, fracturing, or testing, another 12 wells in the process of being drilled, and another 73 wells permitted but not yet drilled (LNDR, 2012a). As of December, 2012, there were 784 producing Haynesville wells in Texas. Natural gas production from the Haynesville shale formation surpassed Texas’s Barnett Shale in February, 2011 (Figure 1-2) making the Haynesville the top-producing U.S. natural gas play in 2011 (U.S. Energy Information Administration (EIA, 2011a).

Figure 1-2. Haynesville and Barnett Shale gas production (bcf/day) (EIA, 2011b).

January 2013

12

1.1 Shale Activity in the U.S.

Active shale plays in the U.S. include the Barnett, Haynesville, Eagle Ford, Fayetteville, Woodford, Antrim, Marcellus, Niobrara, and Bakken Shales. These shale formations are spread over a very wide area, stretching from the Northeast, through the South and Gulf States, the Midwest, and the Rocky Mountain region. Figure 1-3 is a map of active shale plays in the continental U.S. (lower 48 states).

Some of these shale plays have been actively developed and are highly productive, while others are in earlier stages of development. They are primarily gas plays, but notable active oil shale plays are the Eagle Ford and Bakken Shale formations. As shown in Figure 1-3, many of these shale formations have a large geographic extent, including multiple counties in multiple states. Table 1-1 summarizes the production of shale gas in the U.S. in the period 2007-2010. Figure 1-4 shows, the U.S shale gas production per day from 2000 to 2012.

Table 1-1. U.S. shale gas production 2007-2010. 2007 2008 2009 2010

Shale Gas (billion cubic feet)

1,239 2,116 3,110 5,336

Figure 1-3. Major U.S. shale plays in the lower 48 states (EIA, 2011c).

January 2013

13

Figure 1-4. U.S Shale Gas Production (bcf/day) from 2000 to 2012 (EIA, 2012).

1.2 The Haynesville Shale

Development of the Haynesville began in 2007-2008. The first wells drilled were highly productive and sparked intensive leasing and exploration activity. NETAC determined that although the development of natural gas resources in the Haynesville was likely to be economically important, it could also generate significant emissions of ozone precursors in a region with several ozone monitors that were close to exceeding the 2008 National Ambient Air Quality Standard for ozone. During spring 2009, NETAC developed an emission inventory of ozone precursors for projected future Haynesville Shale development from 2009 through 2020 (ENVIRON, 2009) and evaluated the impact of the Haynesville emissions on ozone in Northeast Texas (Kemball-Cook et al., 2010).

1.2.1 Previous Haynesville Emission Inventory Study

Developing natural gas resources in the Haynesville Shale requires significant exploration (i.e. drilling, completion) activities and, as the field develops, construction of production and gathering/transmission infrastructure. Each of these activities results in a significant population of equipment operating in the region, potentially contributing to air emissions of ozone precursors. Given the existing and potential levels of activity in the Haynesville Shale, the ongoing development of the formation could result in significant emissions and impacts to local air quality.

January 2013

14

In 2009, ENVIRON developed the first Haynesville Shale emission inventory estimates. The emission inventory showed the potential ozone precursor emission over short- and long-term time periods for three development scenarios. Base year 2009 emissions were developed using a bottom-up methodology and three future emission scenarios were developed for the projections of the inventory to years 2010-2020.

Objective

The purpose of the previous Haynesville Shale emissions inventory presented in ENVIRON (2009) was to quantify emissions of ozone precursors from the most significant source categories resulting from developing the Haynesville Shale formation.

Methods

Based on previous relevant studies including Central Regional Air Planning Association (CENRAP) state oil and gas emission inventories (Bar-Ilan, et al. 2008b) and a study documenting the development of an emissions inventory for the Barnett Shale (Armendariz, 2009), emissions for the ENVIRON (2009) study were estimated for drill rigs, fracing, completion venting, blowdown venting, fugitives, pneumatic devices, heaters, dehydrators, flaring, wellhead compressors and midstream compressor stations and gas plants.

The 2009 survey questionnaire was distributed to major Haynesville producers asking for representative information on configurations of equipment, equipment specifications and usage, and process data on venting, dehydration and other processes. However, no operator provided survey responses that could be used to refine emissions estimation inputs. The exploration and production activity data were mainly obtained from Bar-Ilan, et al. (2008b) and Armendariz (2009) whereas for midstream sources, emission rate estimates were based on 2004 emissions obtained from the Louisiana Department of Environmental Quality (LADEQ) and the Texas Commission on Environmental Quality (TCEQ) scaled to future year Haynesville Shale formation natural gas production.

The base year for the ENVIRON (2009) study was 2009 as this was the first year for which sufficient production data from wells drilled in the Haynesville Shale were available for well decline curve and future year natural gas production projections. Emissions were estimated for the ozone precursors nitrogen oxides (NOx), volatile organic compounds (VOC) and carbon monoxide (CO). A bottom-up approach was used to estimate the base year emissions inventory. The methodology for developing the emissions inventory was based on that used in the Western Regional Air Partnership (WRAP) Phase III studies (Bar-Ilan et al., 2008a, Bar-Ilan, et al., 2009a, Bar-Ilan, et al., 2009b, Bar-Ilan, et al., 2009c, Bar-Ilan, et al., 2009d, Bar-Ilan, et al., 2010, Bar-Ilan, et al., 2011, Bar-Ilan, et al., 2012) and CENRAP oil and gas emission inventories (Bar-Ilan et al., 2008). Projections of future activity and emissions were developed for the Haynesville Shale for each year from 2010 to 2020. Exploration and production data from the Railroad Commission of Texas (TRRC) and the Louisiana Department of Natural Resources (LDNR) were used to project future drilling and gas production activity. These projections were used to scale the base year emissions inventory to all future years, using activity factors (number of new wells drilled, gas production, and condensate production estimates) as

January 2013

15

projection surrogates. The study evaluated three scenarios corresponding to limited, moderate, and aggressive development for years 2009-2020 that were based on projected annual growth of drilling rigs in the Haynesville Shale, with emissions estimates developed for each scenario.

Limitations

At the time of the ENVIRON (2009) study, development of the Haynesville Shale was in its initial stages. Basic information, such as the geographic extent and recoverable reserves of the Haynesville Shale, were uncertain. Active Haynesville Shale wells had been producing gas for a very limited period of time; therefore long-term yearly production rates were unknown. Producers and regulatory agency personnel indicated that only dry gas with no condensate was expected to be produced from these wells, but no detailed gas compositions were available for use in the inventory development. Northern California gas composition data was chosen because, like the Haynesville Shale formation, fields in Northern California produce dry natural gas exclusively. Other Texas and Louisiana formations were considered for use as proxies for Haynesville Shale natural gas composition, but all formations for which ENVIRON was able to find natural gas compositions produced significant amounts of condensate.

As part of the ENVIRON (2009) study, a survey was distributed to Haynesville Shale operators. The survey queried the operators on their equipment operating in the field, gas composition, and their projected future activity in the Haynesville. Because so few wells had been drilled in the Haynesville Shale at that time, several producers felt that they did not yet have enough information to predict future year activity and production, and all of the producers declined to participate in the survey. Therefore, one limitation of the earlier study was the reliance on equipment, process and emissions factor data from broader regional studies such as the

CENRAP (2008) and WRAP Phase III studies (Bar‐Ilan et al., 2008a, Bar‐Ilan, et al., 2009a,

Bar‐Ilan, et al., 2009b, Bar‐Ilan, et al., 2009c, Bar‐Ilan, et al., 2009d, Bar‐Ilan, et al., 2010,

Bar‐Ilan, et al., 2011, Bar‐Ilan, et al., 2012).

Ozone Impacts

NETAC’s photochemical modeling of the year 2012 using the ENVIRON (2009) Haynesville Shale emission inventory showed that 8-hour ozone impacts occurred within Northeast Texas and Northwest Louisiana as a result of development in the Haynesville Shale with projected design value increases up to 5 ppb at area ozone monitors for the aggressive development scenario (Kemball-Cook et al., 2010). The ozone modeling results showed 2012 design value increases for ozone monitors located within the Louisiana Haynesville Shale counties of Bossier and Caddo of 2 ppb in the low scenario and 4-5 ppb in the high scenario. The Karnack monitor also showed design value increases of 2 ppb in the low scenario and 4-5 ppb in the high scenario. For the Longview and Tyler monitors, which were located west of the Haynesville Shale at the time of the 2009 study, design value increases were smaller, ranging from 1 ppb for both monitors in the low scenario to 1-2 ppb in the high scenario. Ozone increases due to Haynesville Shale emissions also affected regions outside Northeast Texas and Northwest Louisiana due to ozone transport. The photochemical modeling results indicated that Haynesville Shale development is an area of concern for future air quality in Northeast Texas.

January 2013

16

1.2.2 Haynesville Development Since the Previous Haynesville Shale Inventory

In early 2009, when the original Haynesville inventory was compiled, there were 95 drilling rigs active in the Haynesville. By 2010, the number of drilling rigs active in the Haynesville increased dramatically and peaked at approximately 180 rigs (Figure 1-5), which was well ahead of the pace of drilling predicted in the ENVIRON (2009) study. The projected number of drill rigs in operation in the Haynesville Shale for the high, medium and low emissions scenarios for the period 2009-2020 is shown in Figure 1-6. The blue bars in Figure 1-6 show snapshots of actual rig counts in March of each year. In March 2010, there were over 160 rigs drilling in the Haynesville; this far exceeded the high scenario prediction. Since 2010, pace of development has slowed due to low natural gas prices and high ratio of oil-to-gas prices (Figure 1-5). In March 2011, the number of active rigs in the Haynesville was close to the predicted value for the aggressive scenario, but by March 2012, the number of rigs operating in the Haynesville had fallen sharply. As of December 2012, there are 24 rigs active in the Haynesville; this figure is far below the low scenario prediction of 95 rigs.

Figure 1-5. Count of drill rigs active in the Haynesville Shale from January 2010-December 2012. Figure from http://www.haynesvilleplay.com/.

January 2013

17

Figure 1-6. Projected annual average drill rig count for low, medium and high scenarios from ENVIRON (2009). Blue bars show actual drill rig counts for March of each year from 2009-2012. Drill rig counts from http://www.haynesvilleplay.com/.

The most recent Haynesville data from the Louisiana Department of Natural Resources (December 2012; LDNR, [2012c]) and the Texas Railroad Commission (December 2012; TRRC, [2012]) indicate that there were approximately 784 producing Haynesville Shale wells in Texas and 2,141 producing wells in Louisiana for a total of 2,925 producing Haynesville wells. ENVIRON (2009) forecast 1,568, 1,875, and 2,181 actively producing wells at the end of 2012 for the low, moderate, and aggressive scenarios, respectively. The December 2012 well count indicates that the number of active wells predicted by even the aggressive scenario for the end of 2012 has been exceeded. In addition, Haynesville Shale development has grown well beyond its 2009 boundaries into Gregg, Marion, San Augustine, Angelina, and Sabine Counties in Texas and Webster parish in Louisiana as exploration expanded the boundary of the known Haynesville formation.

Given the large number of active, producing wells, the Haynesville Shale continues to be an emissions source that must be monitored and accurately represented in NETAC’s ozone modeling. The comparison between the predicted and actual drill rig activity and well count as well as the expanded spatial extent of the Haynesville Shale (Figure 3-1) indicate that the future year projections made in the ENVIRON (2009) study are now out of date.

January 2013

18

2.0 OBJECTIVE OF THIS STUDY

The previous ENVIRON (2009) study that assessed emissions and air quality impacts from the Haynesville Shale was limited by data availability and the early state of development of the Haynesville Shale at the time of the study. NETAC therefore has undertaken a new study to update and improve the inventory and subsequent air quality impact assessment as described in this report. The primary objectives of this new study are:

1. To improve the original Haynesville Shale emission inventory using detailed equipment,

process, well configuration, emissions factor, and usage data about Haynesville wells

obtained from a survey of Haynesville operators;

2. To incorporate current Haynesville Shale gas production and well statistics that reflect the

recent changes in the rate of development in the Haynesville;

3. To update the ENVIRON (2009) future year projection scenarios to incorporate more recent

data on the Haynesville Shale

The emission inventory results will be used to produce model-ready emission inventory for NETAC’s ozone modeling, and this modeling will be conducted and reported upon in a separate report.

January 2013

19

3.0 SCOPE OF STUDY

The scope of the updated Haynesville Shale emissions study is discussed below, including the pollutants addressed, the geographic scope, and the temporal scope.

3.1 Pollutants

The most important air quality issue for Northeast Texas is ozone. Therefore, the focus of the emission inventory is on the major ozone precursors; air emissions of NOx, VOC, and CO are estimated.

3.2 Geographic Extent of the Haynesville Shale

As in the original Haynesville study (ENVIRON, 2009), this study is restricted to Haynesville Shale development in Northeast Texas and Northwest Louisiana. Both the TRRC and the LDNR maintain Haynesville Shale-specific websites that summarize the information gathered by these agencies on the Haynesville Shale and track the progress of its development (TRRC, 2012, LDNR, 2012c). Both the TRRC and LDNR have published maps delineating the geographic extent of the Haynesville Shale based on exploration to date. Figure 3-1 shows the Haynesville Shale regional maps developed by the TRRC for Texas and by the LDNR for Louisiana. The left hand panels of Figure 3-1 show the extent of Haynesville development at the time of the original ENVIRON (2009) inventory, and the right hand panels show development as of December 2012.

Significant oil and gas exploration and production in formations unrelated to the Haynesville Shale are present in both Northeast Texas and Northwest Louisiana. The identification of specific wells that access the Haynesville Shale formation was made by the state agencies through a combination of information that includes the drilling depth, the composition of the gas produced, and, if available, core samples indicating that a well has accessed the Haynesville Shale. The Haynesville Shale is defined to be the stratigraphic interval consisting of shale below the base of the deepest Cotton Valley sands and above the top of the Smackover Formation. It was beyond the scope of this effort to evaluate these criteria or develop any additional analysis to determine the extent of the Haynesville, and the definitions presented by the TRRC and LDNR were used directly.

As of December, 2012, the TRRC had identified four core and six non-core counties in Texas as Haynesville/Bossier counties, and the LDNR reported eight parishes with Haynesville Shale drilling activity in Louisiana. For the purposes of this study, these 18 counties/parishes were defined to be Haynesville Shale region counties and parishes. The counties/parishes that comprise the geographic scope of this analysis are listed below in Table 3-1.

January 2013

20

February, 2009 December 2012

Figure 3-1. Upper left panel: Spatial extent of the Haynesville Shale in Texas by county and well locations as of February, 2009. Upper right panel: Spatial extent of the Haynesville Shale in Texas by county and well locations as of December, 2012. Figures from the Railroad Commission of Texas website (TRRC, 2012). Pale blue shading indicates that the TRRC considers that county to be a core Haynesville County. Yellow shading indicates that the TRRC considers that county to be a non-core Haynesville County. Lower left panel: Extent of the Haynesville Shale in Louisiana as of February, 2009 (LDNR, 2012c). Lower right panel: Extent of the Haynesville Shale in Louisiana as of December, 2012 (LDNR, 2012c).

http://www.rrc.state.tx.us/bossierplay/index.php

January 2013

21

Table 3-1. Counties in Texas and parishes in Louisiana considered part of the Haynesville Shale development area.

Texas Counties Louisiana Parishes

Harrison (core) Caddo

Panola (core) Bossier

Shelby (core) Webster

San Augustine (core) Bienville

Rusk (non-core) De Soto

Nacogdoches (non-core) Red River

Angelina (non-core) Natchitoches

Gregg (non-core) Sabine

Marion (non-core)

Sabine (non-core)

Figure 3-2 below shows the extent of the Haynesville Shale including counties and parishes in both Texas and Louisiana. These counties were considered in this study to form the geographic extent of the Haynesville Shale for exploration and well site production activities. All source categories associated with the development and operation of well sites were assumed to occur within these counties for all future years.

January 2013

22

Figure 3-2. Spatial extent of the Haynesville Shale in Texas and Louisiana as defined in this study.

An expanded geographic domain was considered for emissions sources associated with gas transmission and processing, specifically central compressor stations and gas processing facilities. These facilities are often not owned and operated by the same companies that own and operate the wells and produce the gas, and are collectively referred to in this analysis as “midstream” facilities. It was expected that gas produced in the Haynesville Shale region would be gathered and processed at nearby midstream facilities, but pipelines could carry produced gas to facilities located outside of the counties identified as the Haynesville Shale development area in this study. This is particularly true for gas processing plants which may not be located directly within a development area.

The locations of midstream facilities operating within the Haynesville Shale and surrounding regions were extracted from emission inventory databases maintained separately by the Texas Commission on Environmental Quality (TCEQ) and the Louisiana Department of Environmental Quality (LADEQ). Similar to the previous ENVIRON (2009) study, a subset of the midstream sources were identified which lie within or near the periphery of the Haynesville Shale region. Figure 3-3 shows a map of midstream facility locations and the boundaries of the region in

January 2013

23

which midstream facilities were considered. Table 3-2 lists the counties and parishes for which emissions from midstream sources were used in the inventory development.

Table 3-2. Counties in Texas and parishes in Louisiana considered part of the Haynesville Shale midstream gas gathering and processing area.

Texas Counties Louisiana Parishes Anderson Bienville

Camp Bossier

Cass Caddo

Cherokee Claiborne

Franklin De Soto

Gregg Jackson

Harrison Lincoln

Henderson Natchitoches

Hopkins Red River

Marion Sabine

Morris Webster

Nacogdoches Winn

Panola

Rains

Rusk

Sabine

San Augustine

Shelby

Smith

Titus

Upshur

Van Zandt

Wood

January 2013

24

Figure 3-3. Midstream facilities in Texas and Louisiana (identified from state emission inventory databases) in the Haynesville Shale and surrounding region considered in this analysis.

A more detailed description of the procedure used to estimate midstream facility emissions is provided in the Section 3 on the emissions inventory development methodology.

3.3 Temporal Scope

The study considers a base year of 2011 with future year emissions projected annually for the period 2012-2020. The latest year for which complete production statistics were available from the TRRC and LDNR was 2011. Therefore, for purposes of comparison with the previous study and for purposes of evaluating predicted future impacts, emissions were estimated for each year in the period 2011-2020.

January 2013

25

4.0 METHODOLOGY

This section describes the emissions inventory for the Haynesville Shale for the three future year scenarios considered in this analysis: (1) low development; (2) moderate development; and (3) aggressive development. The emissions inventory considers a number of gas production source categories which are described below. The description of each source category provides a detailed description of the methodology used to estimate emissions on a unit level, the scale-up of these emissions to the entire Haynesville Shale region, and lists the sources of data used in the inventory calculations.

The main objective of the study is to improve the existing Haynesville Shale emission inventory (ENVIRON, 2009). For this effort, a survey of Haynesville Shale oil and gas operators was conducted to gather information on several key source categories. The details of the survey outreach process and results are described in this section. This section also describes how the base year Haynesville Shale production for Northeast Texas and Northwest Louisiana are derived along with the production forecast methodology.

4.1 Survey

Previous regional natural gas emission inventory work has shown that the characteristics of exploration and production equipment, well configuration, usage, and processes can vary significantly with geographic location and formation; therefore, incorporation of Haynesville-specific survey data would increase the accuracy of the emission inventory. In May 2012, a short survey was sent out to the operators active in the Haynesville Shale to collect Haynesville-specific data for use in developing the emission inventory.

4.1.1 Survey Outreach Process and Results

The survey effort focused on several key emissions source categories based on the findings of the 2009 Haynesville Shale emission inventory. The survey requested information for the following source categories:

Wellhead Compressor Engines

Drilling Rigs

Condensate Tanks

Well Pad Construction

Heaters/Heater-Treaters

Additionally, the survey requested that the operators provide a representative Haynesville Shale produced gas composition for use in calculating emissions for sources that release natural gas into the atmosphere. The survey and cover letter that accompanied its transmittal to the operators are shown in Appendix A. The companies contacted were Anadarko Petroleum Corporation, BP America, Chesapeake Energy Corp., Devon Energy, EOG resources, Exco Resources Inc., GMX Resources, Goodrich Petroleum company, NFR Energy LLC, Penn Virginia Oil and Gas LP, Samson Lonestar and XTO Energy Inc. These operators were selected based on

January 2013

26

TRRC natural gas production statistics as being the top producers in the Haynesville Shale in 2011; it was assumed that the survey responses from these operators would be broadly representative of all activity in the Haynesville. Table 4-1 shows production ownership by the surveyed producers.

Table 4-1. 2011 Haynesville Shale gas production ownership by surveyed producers.

Surveyed Operators

Total Gas Production in

Haynesville Region Percentage of Total

Haynesville Production

Survey Received

(BCF/yr) (Y/N)

Anadarko Petroleum Corporation 13.6 0.6% N

BP America 2.7 0.1% N

Chesapeake Energy Corp. 576.5 23.6% Y

Devon Energy 14.2 0.6% N

EOG resources 111.2 4.5% N

Exco Resources Inc. 400.8 16.4% N

GMX Resources 22.6 0.9% N

Goodrich Petroleum Company 21.8 0.9% N

NFR Energy LLC 26.7 1.1% N

Penn Virginia Oil and Gas LP 5.8 0.2% Y

Samson Lonestar 38.6 1.6% N

XTO Energy Inc 153.1 6.3% Y

Surveyed Operator Total 1,387.40 56.8%

Responding Operator Total 735.40 30.1%

Haynesville Shale Formation-wide Total 2,444.20 100.0%

Of the twelve companies contacted, three provided responses. These three companies, Chesapeake Energy, Penn Virginia Oil and Gas, and XTO Energy, represent 30% of the total Haynesville production for 2011. The responses from these companies are deemed to be the best and most Haynesville-specific information available and, for the purposes of emission inventory development, are assumed to be representative of Haynesville Shale formation-wide oil and gas operations.

For emissions from those source categories that require estimates of volume of gas vented or leaked, such as well blowdowns, completions, pneumatic devices and fugitive emissions, gas composition analyses were requested from all participating companies for gas produced from the Haynesville formation. These composition analyses were averaged to derive a representative produced gas composition. The average composition analysis was used to determine the average VOC volume and mass fractions of the vented gas.

January 2013

27

For all equipment source categories, operators were asked to provide basic information including typical equipment and usage characteristics on a per well basis. The questionnaire asked operators to provide information on hours of operation, horsepower, fuel consumption and equipment Tier level for drill rigs used to drill their Haynesville wells. Similarly, the hours of operation, typical equipment configuration, load factor, number of equipment, and firing rate were requested for the well pad construction equipment, wellhead compressor engines and heaters as applicable. The well pad area of a typical Haynesville well was also requested. In addition, operators were requested to provide an estimate of the fraction of wells with wellhead compressor engines and the prevalence of engine emissions controls. Typical configurations for condensate and produced water storage tanks at well sites were requested. The survey also asked operators to provide a flash gas composition analysis and any supplemental documents to support the emissions calculations including input/output model runs for production equipment modeling software such as E&P Tank, HYSYS, or Promax if these were available.

Survey responses were reviewed to determine if specific data were beyond the typical range of values for a parameter compared to data from other oil and gas emission inventory efforts. Data that were determined to be beyond a typical range without sufficient rationale were rejected from inclusion in the final inputs compilation. For each quantity in the survey responses, the appropriate surrogate (well count, gas production, or condensate production) was used to derive a representative value for the entire Haynesville Shale.

The representative data derived from the survey responses were compared with the data used in the previous Haynesville Shale emission inventory.

Drilling Time: The updated estimates of drilling time and horsepower from survey data were significantly lower, approximately 64% and 69% respectively, than the data used in previous emission inventory.

Drill Rig Engine Tier: The previous emission inventory used uncontrolled drill rig emission rates consistent with Bar-Ilan, et al (2008b), however, in this study the operators have indicated that cleaner technology Tier II engines are used for drilling a typical Haynesville well; this results in lower emissions rates relative to the previous emission inventory.

Well Pad Construction: At the time of the original emission inventory development, data was unavailable for the well pad construction equipment and hence emissions from this category were not included. For the present study, the operators were able to provide enough information to allow the calculation of emissions from this category.

Well Site Heaters: The updated survey data resulted in a lower heater firing rate and a lower number of hours of operation and number of heaters per well compared to the previous emission inventory.

Well Head Compression: The survey indicated that operators only use wellhead compressor engines for a very small fraction of wells (approximately 0.08%) compared to the previous emission inventory assumption of 2% of wells with wellhead compressor engines. Although two operators provided wellhead engine horsepower information, the sample size was

January 2013

28

small enough that it was determined that using engine data from the TCEQ’s Barnett Shale Phase I emission inventory would allow our estimates to be based on a much larger engine population. It was assumed that the Barnett engine population is similar to that of the Haynesville. Therefore, the weighted average horsepower and fraction of rich and lean burn engines were obtained from the Barnett Shale Phase I emission inventory (TCEQ 2011a). The final horsepower value used was 127 horsepower compared to 242 horsepower used in the previous (ENVIRON, 2009) emission inventory. The TCEQ has developed a Phase II inventory of the Barnett Shale but detailed data on compressor engine population distribution by horsepower were not available for the Phase II inventory at the time of this study, therefore data was taken from the Phase I inventory.

Produced Gas Composition Analysis: Operators were able to provide the Haynesville regional specific gas composition analyses which are compared with the original emission inventory values in Table 4-2.

Table 4-2. Comparison of gas composition analyses used in original and updated emission inventories.

Natural Gas Properties Previous Study

(ENVIRON, 2009) Updated Estimates

Heating Value of Gas (BTU/SCF) 950 950

Gas Molecular Weight 17.2 17.0

VOC Fraction (molar) 0.43% 0.34%

VOC Molecular Weight 58.9 52.3

Component Weight Fractions Weight

Fractions

CO2 0.3% 6.07%

N2 11.8% 0.43%

Methane C1 85.4% 90.16%

Ethane C2 1.1% 2.29%

Propane C3 0.5% 0.52%

i-Butane i-C4 0.2% 0.20%

n-Butane n-C4 0.2% 0.13%

i-Pentane iC5 0.2% 0.07%

n-Pentane nC5 0.1% 0.03%

Hexanes C6 0.4% 0.10%

4.2 PRODUCTION STATISTICS DEVELOPMENT

This section discusses the base year (2011) production data compilation methodology. The base year production data for Texas is from the TRRC and Louisiana production data is from the Louisiana Department of Natural Resources (LDNR).

4.2.1 Drilling (Spuds) Estimates

The number of Haynesville Shale spuds drilled in 2011 was estimated based on the number of rigs operating in the Haynesville Shale and the average time required to drill a well. For Northwest Louisiana, the scout report (LNDR, 2012b) provides spud data by well API number, by field, by parish, by operator name, by status of wells and by year. This data could have been

http://www.tceq.texas.gov/airquality/point-source-ei/psei.html#barnett

January 2013

29

used to determine the number of new wells drilled in Haynesville region in 2011 for Northwest Louisiana. However due to the complexity involved in determining the number of wells drilled in 2011 for Northeast Texas from the RCC drilling database, which allows users to query drilling permit by field, by permit submission or approve date and by county, drill rig activity was used to determine the number of wells drilled per year in the Haynesville Shale region.

The detailed rig count by formation from February 2011 to December 2011 was obtained from the Baker Hughes database (Baker Hughes, 2012). Data from January 2011 were not available at the time the study was performed. The number of wells drilled per year by a rig operating in the Haynesville Shale was estimated to be 15.6 based on the average operator-provided estimate of 22.3 days of drill rig operation per well, assuming one day for rig moving between drilling events. A 2008 success factor of 55% was assumed to increase linearly to 100% in 2018 as in ENVIRON (2009).

4.2.2 Well Counts and Production Estimates

As explained above, the scout report (LNDR, 2012b) provides spud data by well API number, by field, by parish, by operator name, by status of wells and by year. Based on the year, parish and well production status, the active well counts were estimated for the Louisiana Haynesville region. Oil and gas production statistics for Louisiana Haynesville wells were obtained from the LNDR SONRIS database (LDNR, 2012d) which included gas production and oil production for the year 2011. Raw data was organized by lease unit well (LUW) code and by operator. The database contained duplicate entries from instances where multiple leases were assigned to a single well-pad, potentially leading to double counting of the production for these particular leases. These duplicates were removed from the database by selecting the first unique entry in the database and deleting the following repeated entry in order to compile total production statistics by operator for the Louisiana Haynesville Shale region.

The TRRC maintains a production database which allows users to query production data by parameters such as county, lease name, and field name. The Northeast Texas Haynesville production data were obtained from the TRRC website for the fields containing Haynesville or Bossier in the field name and, in 2011, included the following fields: Carthage (Haynesville Shale), Gladewater (Haynesville), Waskom (Haynesville), and Manning (Bossier Sand). For each field, the gas and condensate production data were queried by lease and county name for calendar years 2008-2011.

January 2013

30

Table 4-3 summarizes the Northeast Texas and Northwest Louisiana active well counts data for the Haynesville region for calendar year 2008-2011. The base year annual gas production and condensate production data for the Haynesville region is shown in Table 4-4.

January 2013

31

Table 4-3. Summary of Northeast Texas and Northwest Louisiana Active Well Counts for Years 2008-2011.

Year Texas -Active

Wells Louisiana -

Active Wells Total Well Counts New Well Counts

2008 97 55 152 152

2009 235 393 628 477

2010 619 1061 1,680 1,051

2011 658 1843 2,501 821

Table 4-4. Summary of Base Year Gas and Condensate Production for the Haynesville Region.

Production Texas - Haynesville

Region Louisiana - Haynesville

Region Total

Gas Production (MCF) 433,549,245 2,010,615,206 2,444,164,451

Condensate Production (bbl) 54,457 9,012 63,469

4.3 Permitted Point Sources

Emissions are generated from midstream sources transporting and processing natural gas produced from the Haynesville Shale formation. Two major source categories are compressor stations and natural gas processing plants. Pipelines may be in the process of being constructed for the purpose of transporting the Haynesville Shale natural gas to markets but it is beyond the scope of this study to determine emissions due to their construction.

At the time the emissions inventory was compiled (mid-late 2012), 2010 was the latest year for which complete permitted source emissions were available for Texas and Louisiana. Therefore, all point source emissions are for the calendar year 2010. Permitted source emissions were obtained for Louisiana for the year 2010 from the State Implementation Plan (SIP) modeling emission inventory being developed by ENVIRON. For point sources with emissions available from EPA's Acid Rain Program/Clean Air Markets Division (ARP/CAMD) database, NOx emissions were taken from the ARP/CAMD database and VOC and CO emissions for these sources were taken from the LADEQ point source database. For sources not in the ARP/CAMD database, emissions were taken from the LADEQ point sources database. Facilities with North American Industry Classification System (NAICS) codes 211111 (Crude Petroleum and Natural Gas Extraction), 211112 (Natural Gas Liquid Extraction), and 486210 (Pipeline Transportation of Natural Gas) were assumed to be midstream sources.

Permitted sources were obtained for the year 2010 for Texas from emissions totals by facility provided by TCEQ (Boyer, 2012). Oil and gas midstream sources were identified for inclusion in the Haynesville Shale emission inventory based on the assumption that facilities with the following standard industrial classification codes (SIC) are midstream facilities: 13*, 492*, and 4612.

January 2013

32

To incorporate midstream emissions for the Haynesville Shale formation, the 2010 Haynesville Shale region midstream emissions are scaled by the ratio of Haynesville Shale formation produced natural gas to 2010 produced natural gas in the Haynesville Shale midstream region (see Table 3-2). The validity of this method as a reasonable estimation is based on assumptions that 2010 Haynesville Shale region natural gas was processed relatively locally and that Haynesville Shale region natural gas was not shipped elsewhere for processing. It is also assumed that natural gas losses in the 2010 Hayesville region are fractionally equivalent to predicted Haynesville Shale formation losses. The major midstream point sources in the Haynesville Shale region and surrounding areas are shown in Figure 3-3.

Total 2010 natural gas production from the counties shown in Table 3-2 was 2,623 billion cubic feet, as summarized in Table 4-5 below.

Table 4-5. 2010 gas production in all Texas counties and Louisiana parishes considered part of the Haynesville Shale midstream/gas processing region.

Texas County Gas Production

(MCF) Louisiana Parish Gas Production

(MCF) Anderson 4,996,766 Bienville 127,860,480

Camp 328,315 Bossier 157,206,494

Cass 743,253 Caddo 224,604,370

Cherokee 25,494,865 Claiborne 17,575,636

Franklin 1,321,795 De Soto 644,455,819

Gregg 46,751,093 Jackson 43,576,835

Harrison 173,241,943 Lincoln 35,705,071

Henderson 18,286,865 Natchitoches 90,309

Hopkins 184,321 Red River 248,222,769

Marion 3,355,121 Sabine 41,853,383

Morris 0 Webster 45,220,854

Nacogdoches 121,644,458 Winn 380,223

Panola 293,984,195

Rains 2,031,875

Rusk 125,066,891

Sabine 304,835

San Augustine 67,523,887

Shelby 78,942,515

Smith 27,924,337

Titus 0

Upshur 36,168,445

Van Zandt 2,925,540

Wood 4,956,255

TOTAL 1,036,177,570 1,586,752,243 Note: 2010 gas production was used to derive emission factors (tons/mmscf) for point sources. The methodology used to calculate emissions factors is described below in Section 4.4.14.

January 2013

33

4.4 Methodology for Base Year Emissions

The purpose of the Haynesville Shale emissions inventory is to quantify emissions of ozone precursors from the most significant source categories resulting from the development of the Haynesville Shale formation. Emissions were estimated for the major contributing sources listed in Table 4-6. Source categories were grouped as either “exploration and production” or “midstream/gas processing” sources. The exploration and production grouping generally refers to source categories that would be located at the well site and would be owned and operated by production companies. The midstream/gas processing grouping refers primarily to compressor stations and major gas processing facilities that would be located downstream of the well sites and are typically owned and operated by midstream companies (but in the case of compressor stations may be owned by production companies).

Table 4-6. Haynesville Shale emission inventory source categories. Phase Source Category

Exploration and Production Wellpad Construction

Drill Rigs

Fracing

Completion Venting

Blowdown Venting

Fugitives

Pneumatic device

Heaters

Dehydrators

Flaring

Wellhead Compressors

Condensate Tanks

Midstream/Gas Processing Compressor Stations and Gas Processing Plants

For source categories for which operators provided Haynesville Shale specific information, this survey data was used directly. In the absence of Haynesville-specific data from operators, ENVIRON referred to other oil and gas emission inventory studies and adapted a methodology that could be used to estimate Haynesville Shale emissions. For exploration and production sources, emission rates were estimated based on data gathered primarily from Bar-Ilan, et al. 2008b and Armendariz, 2009. For midstream sources, emission rate estimates were based on 2010 emissions obtained from the LADEQ and 2010 emissions obtained from TCEQ scaled to future year Haynesville Shale formation production.

The methodologies for developing unit-level emissions factors for base year (2011) are described below and the quantity used to scale the unit level emissions up to the formation-wide emissions for the entire Haynesville Shale are listed in Table 4-7.

January 2013

34

Table 4-7. By source category scaling parameter. Category Scaling Parameter

Well Pad Construction Spuds

Drill Rigs Spuds

Fracing Spuds

Completion Venting Spuds

Blowdown Venting Active Well Count

Fugitives Active Well Count

Pneumatic device Active Well Count

Heaters Active Well Count

Dehydrators Gas Production

Flaring Gas Production/Condensate Production

Wellhead Compressors Active Well Count

Condensate Tanks Condensate Production

Midstream Sources (Compressor Stations and Gas Plants) Gas Production

4.4.1 Well Pad Construction

Methodology

The well pad is the plot of land that is designated for natural gas or oil extraction. The first step in well development is to clear the pad area in preparation for the drilling of the well. The pad is constructed by clearing all trees and obstacles to allow for engineering of a foundation usually made of local rock from a quarry.

Emissions from a single well pad construction are determined according to Equation 1:

185,907,

wellpadaveragetotali

enginewellpad

tLFHPEFE

Equation (1)

where: Ewellpad,engine is the emissions from a well pad construction for one well [ton/spud] EFi is the emissions factor for all well pad engines for pollutant i [g/hp-hr] HPtotal is the total horsepower of all engines used for well pad construction [hp] LFaverage is the average load factor for all well pad construction engines tdrilling is the actual on-time of all engines for a typical well [hr/spud]

Extrapolation to Region-Wide Emissions

Well pad construction emissions from a single pad are scaled to formation-wide emissions according to Equation 2:

TOTALenginewellpadTOTALwellpad SEE ,, Equation (2)

January 2013

35

where: Ewellpad,TOTAL is the total emissions in the formation from well pad construction activity [tons/yr] Ewellpad,engine is the total emissions in the formation from single well pad construction [tons/spud] STOTAL is the total number of spuds that occurred in the formation for a given calendar year

Input Data

In the absence of Haynesville Shale-specific data, load factors and emission factors were taken from the EPA NONROAD model (EPA, 2008). Table 4-8 shows the data used to estimate well pad construction emissions.

Table 4-8. Well pad construction emissions estimation inputs. Property Value Source

Dozer

Fuel Type diesel Survey

Rated horsepower 125 Survey

Time used (hrs) 40 Survey

Load Factor 0.21 EPA NONROAD

Emission Factors

(g/bhp-hr)

NOx 9.8 EPA NONROAD Base Engine

VOC 1.0 EPA NONROAD Base Engine

CO 3.7 EPA NONROAD Base Engine

Bull Dozer





Emission Factors

(g/bhp-hr)




Track Hoe





Emission Factors

(g/bhp-hr)




Grader





Emission Factors

(g/bhp-hr)




January 2013

36

4.4.2 Drill Rigs – Drilling Operations

Methodology

Drill rigs are most commonly powered by one or more diesel-fired compression-ignition engines. There are three primary functions of these engines:

1. Draw works – the draw works engine(s) provides power to the rotating drill bit and is

responsible for the actual cutting operation of the rig

2. Mud pumps – the mud pump engine(s) provides pumping of the working fluid (often

referred to as “mud”) into the bore hole for lubrication and cooling as well as pumping the

spent fluid and debris material out of the bore hole

3. Generators – the generator engine(s) provides power to the drilling crew and incidental

power for the entire site operation (lighting, HVAC, crew quarters, etc.), or provides power

to drive the draw works and pump motors in a diesel-electric configuration

Although there are three primary functions of the rig engines, there may be more than one of each engine type with the additional engines either required for additional horsepower or used as back-up engines. Each of these three engine types is used for differing durations throughout a drilling process and is likely to have different load factors. In addition, each of the three engine types is likely to be of differing model years and hence Tier levels, since individual engines on rigs may be replaced on independent turnover schedules.

Emissions from a single drilling event are determined according to Equation 3:

185,907

drillingaveragetotali

drilling

tLFHPEFE

Equation (3)

where: Edrilling,engine is the emissions from a drilling rig for drilling one well [ton/spud] EFi is the emissions factor for all drilling rig engines for pollutant i [g/hp-hr] HPtotal is the total horsepower of all engines on the drilling rig [hp] LFaverage is the average load factor for all engines on the drilling rig tdrilling is the actual on-time of all engines on the drilling rig for a typical drilling event [hr/spud]


Drilling emissions from a single drilling event are scaled to formation-wide emissions according to Equation 4:

TOTALdrillingTOTALdrilling SEE , Equation (4)

where: Edrilling,TOTAL is the total emissions in the formation from drilling activity [tons/yr]

January 2013

37

Edrilling is the total emissions in the formation from single drilling event [tons/drilling event] STOTAL is the total number of spuds that occurred in the formation for a given calendar year

Input Data

Input data was based on the operator survey and Bar-Ilan, et al (2008b) as summarized below in Table 4-9.

Table 4-9. Drill rig emissions estimation inputs. Property Value Source

Fuel Type diesel Bar-Ilan, et al. 2008b- East Texas Basin

Rated horsepower 1,122 Survey

Time used per spud (hrs) 536 Survey

Load Factor 0.67 Bar-Ilan, et al. 2008b - East Texas Basin

Technology Tier II Survey

Emission Factors (g/bhp-hr)

NOx 4.5 EPA Federal Tier Standards

VOC 0.3 EPA Federal Tier Standards

CO 2.6 EPA Federal Tier Standards

4.4.3 Fracing

Methodology

Fracing, short for hydraulic fracturing, is performed after a well bore has been drilled in order to stimulate natural gas production from the well bore. The process of fracing requires the use of pump engines to push large quantities of fluid and sand/glass into the well bore to hydraulically fracture the formation to increase surface area and release natural gas. Generators and other equipment may also be used in this process on an as-needed basis.

Emissions from a single fracing event are determined according to Equation 5:

185,907

fracingaveragetotali

fracing

tLFHPEFE

Equation (5)

where: Efracing is the emissions from fracing engines for one fracing event [ton/fracing event] EFi is the emissions factor for all fracing engines for pollutant i [g/hp-hr] HPtotal is the total horsepower of all engines on the fracing rig [hp] LFaverage is the average load factor for all engines on the fracing rig tdrilling is the actual on-time of all engines on the fracing rig for a typical fracing event [hr/spud]


Fracing emissions from a single fracing event are scaled to formation-wide emissions according to Equation 6:

January 2013

38

TOTALfracingTOTALfracing SEE , Equation (6)

where: Efracing,TOTAL is the total emissions in the formation from fracing activity [tons/yr] Efracing is the total emissions in the formation from single fracing event [tons/fracing event] STOTAL is the total number of spuds that occurred in the formation for a given calendar year

Input Data

In the absence of Haynesville Shale specific data, fracing engine power and activity were taken from a study in which emissions were estimated for the Barnett Shale formation (Armendariz, 2009) and emission factors were taken from the EPA NONROAD model (EPA, 2008) as summarized below in Table 4-10.

Table 4-10. Fracing emissions estimation inputs. Property Value Source

Fuel Type diesel Armendariz, 2009 (Barnett Shale EI)

Rated horsepower 1000 Armendariz, 2009 (Barnett Shale EI)

Time used (hrs) 54 Armendariz, 2009 (Barnett Shale EI)

Load Factor 0.5 Armendariz, 2009 (Barnett Shale EI)

Emission Factors (g/bhp-hr)




4.4.4 Completion Venting

Methodology

Once drilling and other well construction activities are finished, a gas well must be completed in order to begin producing gas. The completion process requires venting of the well for a sustained period of time to remove mud and other solid debris in the well, to remove any inert gas used to stimulate the well (such as CO2 and/or N2) and to bring the gas composition to pipeline grade. During this process, significant amounts of gas may be vented, and this gas can be a VOC emissions source. This analysis assumes that the composition of the completion venting gas is identical to production gas, because no detailed information was available on the composition of completion venting gas.

Emissions from well completions are estimated on the basis of the volume of gas vented during completion and the average VOC content of that gas; the VOC content is obtained from gas composition analyses. Flaring and/or green completion practices may be used to control emissions from the completion process. Flaring typically has a control efficiency greater than 95% for VOC emissions, but generates emissions of NOx. Green completion practices have a range of control efficiencies depending on the amount of vented gas that is captured during the process.

January 2013

39

The calculation methodology for estimating emissions from a single completion event is shown below in Equation 7:

907200

105.3 5,

i

gas

ventedicompletion

f

TMW

R

VPE

Equation (7)

where: Ecompletion,i is the emissions of pollutant i from a single completion event [ton/event] P is atmospheric pressure [1 atm] Vvented is the volume of vented gas per completion [MCF/event] R is the universal gas constant [0.082 L-atm/mol-K] MWgas is the molecular weight of the gas [g/mol] T is the atmospheric temperature [298 K] fi is the mass fraction of pollutant i in the completion venting gas


The total emissions from all completions occurring in a formation are evaluated following Equation 8:

greenflareformationicompletionTOTALcompletion ccSEE 95.01,,

Equation (8)

where: Ecompletion,TOTAL are the total emissions formation-wide from completions [tons/year] Ecompletion,i are the completion emissions from a single completion event [tons/event] cflare is the fraction of completions in the formation that were controlled by flares cgreen is the fraction of completions in the formation that were controlled by green completion techniques Sformation is the formation-wide spud count in the formation for a given calendar year

Input Data

In the absence of Haynesville Shale-specific data, completion venting data were taken from the East Texas Basin data as identified in Bar-Ilan, et al. (2008b).

January 2013

40

Table 4-11 shows the data used to estimate completion venting emissions. Note that the data used in Bar-Ilan et al. (2008b) were for the year 2002, and the fraction of green completions in 2009 is likely to be higher. The emission estimate for completion venting emissions is therefore conservative as the venting emissions will be lower if a higher percentage of completions are carried out using green techniques.

January 2013

41

Table 4-11. Completion venting emissions estimation inputs. Property Value Source

Volume of Gas Vented Per Initial Completion (MCF) uncontrolled 2417

Bar-Ilan et al. 2008b - East Texas Basin

Fraction of completions in the formation controlled by flares 0%


Fraction of completions in the formation controlled by green completion techniques 0%


4.4.5 Blowdown Venting

Methodology

Well blowdowns refer to the practice of venting gas from a well that has developed a cap or obstruction that must be removed before any additional intervention work can be done on the well. Sometimes well blowdowns are conducted on wells that have been shut in for a period of time when the operator desires to bring the well back into production. Well blowdowns are also sometimes conducted to remove fluid caps that have built up in producing gas wells causing a pressure drop across the well orifice that restricts gas flow. Because gas is directly vented from the blowdown event, blowdowns can be a source of VOC emissions.

Emissions from blowdowns are estimated on the basis of the volume of gas vented during a blowdown and the average pollutant content of that gas obtained from gas composition analyses. This methodology is very similar to that of completion venting. Flaring and/or green practices may be used to control emissions from the blowdown process. Flaring typically has a 98% control efficiency for VOC emissions, and green practices have a range of control efficiencies depending on the amount of vented gas that is captured during the process.

The calculation methodology for estimating emissions from a single blowdown event is shown below in Equation 9:

907200

105.3 5

,i

gas

ventediblowdown

f

TMW

R

VPE

Equation (9)

where: Eblowdown,i is the emissions of pollutant i from a single blowdown event [ton/event] P is atmospheric pressure [1 atm] Vvented is the volume of vented gas per blowdown [MCF/event] R is the universal gas constant [0.082 L-atm/mol-K] MWgas is the molecular weight of the gas [g/mol] T is the atmospheric temperature [298 K] fi is the mass fraction of pollutant i in the vented gas

January 2013

42


The total emissions from all blowdowns occurring in a formation are evaluated following Equation 10:

greenflarewellsblowdowniblowdownTOTALblowdown ccNNEE 95.01,, Equation (10)

where: Eblowdown,TOTAL are the total emissions formation-wide from blowdowns [tons/year] Eblowdown,i are the blowdown emissions from a single blowdown event [tons/event] cflare is the fraction of blowdowns in the formation that were controlled by flares cgreen is the fraction of blowdowns in the formation that were controlled by green techniques Nblowdown is the number of blowdowns per well in the formation Nwells is the total number of active wells in the formation for a given calendar year

Input Data

In the absence of Haynesville Shale specific data, blowdown venting data were taken from the East Texas Basin data as identified in Bar-Ilan, et al (2008b). Table 4-12 shows the data used to estimate blowdown venting emissions.

Table 4-12. Blowdown venting emissions estimation inputs. Property Value Source

Blowdown Frequency (events/well/year) 1 Bar-Ilan et al. 2008b - East Texas Basin

Volume of Gas Vented Per Blowdown (MCF) 32 Bar-Ilan et al. 2008b - East Texas Basin

Fraction of blowdowns in the formation controlled by flares 0% Bar-Ilan et al. 2008b - East Texas Basin

Fraction of blowdowns in the formation controlled by green techniques 0% Bar-Ilan et al. 2008b - East Texas Basin

4.4.6 Fugitives (Leaks)

Methodology

Fugitive emissions refer to emissions of produced gas through connectors, flanges, valves and other pipeline hardware at the wellhead. These emissions are essentially leaks that result from high-pressure gas moving through the various hardware components of a wellhead assembly. It should be noted that this source category is distinct from fugitive emissions from pipelines, which are not considered here, and refers only to components located at the wellhead. Because the fugitive emissions are produced gas, this source category can be a source of VOC emissions.

Fugitive emissions from wellheads are estimated using AP-42 emissions factors (EPA, 1995) and component counts for typical well setups. The well setup is typically characterized by the type of equipment installed and by the type of service to which the equipment applies – gas, light liquid, heavy liquid, or water. The typical well setup information for the East Texas Basin was

January 2013

43

developed by Bar-Ilan et al. (2008b). The East Texas basin fugitive component configuration in Bar-Ilan et al. (2008b) includes gas media and light oil media components. The Haynesville Operator survey response indicated that approximately 20% of Haynesville Shale wells are configured to handle both gas and condensate production with the remaining 80% of the wells handling gas, but no condensate production. The typical East Texas Basin fugitive component configuration from Bar-Ilan et al. (2008b), which includes gas media and light oil media components, was applied to 20% of Haynesville Shale wells which are assumed to have both gas and condensate handling infrastructure. The remaining 80% of Haynesville Shale wells are assumed to produce dry gas and have only gas media fugitive components from the East Texas Basin configuration in Bar-Ilan et al. (2008b); the light oil fugitive components are assumed not to be present at these wells. Fugitive emissions for an individual typical well are estimated according to Equation 11:

i

jannualiijfugitive YtNEFE 0011.0,

Equation (11)

where:

Efugitive is the fugitive emissions for a single typical well for pollutant j [ton/yr/well] EFi is the emission factor of TOC for a single component i [kg/hr/component] Ni is the total number of components of type i tannual is the annual number of hours the well is in operation [8760 hr/yr] Yj is the mass fraction of pollutant j to TOC in the vented gas


Formation-wide fugitive emissions are estimated according to Equation 12:

welljfugitiveTOTALfugitive NEE ,, Equation (12)

where: Efugitive,TOTAL is the total fugitive emissions in the formation [ton/yr] Efugitive,j is the fugitive emissions for a single well of pollutant j [ton/yr] Nwell is the total number of active wells in the formation for calendar year 2002

Input Data

In the absence of Haynesville Shale specific data, fugitives data were taken from the East Texas Basin data as identified in Bar-Ilan, et al (2008b).

January 2013

44

Table 4-13 shows the data used to estimate fugitive emissions.

January 2013

45

Table 4-13. Fugitives emissions estimation inputs.

Media Type

Typical Well Equipment Component*

Well without Condensate Gathering

Infrastructure

Well with Condensate Gathering

Infrastructure

Gas

valves 12 12

pump seals 0 0

others 0 0

connectors 35 35

flanges 18 18

open-ended lines 6 6

Light Oil >20º API

valves 0 20

pump seals 0 0

others 0 0

connectors 0 90

flanges 0 0


Heavy Oil <20º API

valves 0 0

pump seals 0 0

others 0 0

connectors 0 0

flanges 0 0


Water/Oil

valves 0 0

pump seals 0 0

others 0 0

connectors 0 0

flanges 0 0


* source: Bar-Ilan et al. 2008b - East Texas Basin

4.4.7 Pneumatic Devices

Methodology

Pneumatic devices are those devices used for a variety of wellhead processes which are powered mechanically by high-pressure produced gas as the working fluid – i.e. pneumatically-powered devices. This is necessary for many remote well sites where electrical grid power is not available to power these devices. Typical pneumatic devices include pressure transducers, liquid level controllers, pressure controllers and positioners. These devices are typically in operation continuously throughout the year. All of these devices vent the working fluid, which is produced gas, and are therefore a source of VOC emissions. Like fugitive emissions, the emissions from these devices are typically estimated by obtaining a configuration of a typical

January 2013

46

well, including the count of devices by type at the typical well. Emissions rates of gas from these pneumatic devices have been studied extensively by the EPA as part of the Natural Gas Star program (EPA, 2004), which are the source of quantitative emissions factors for pneumatic devices in this analysis.

The methodology for estimating the emissions from pneumatic devices for a single typical well is shown in Equation 13:

5

,

105.3907200

TMW

R

PtNV

fE

gas

i

annualii

j

jpneumatic

Equation (13)

where: Epneumatic,j is the total emissions of pollutant j from all pneumatic devices for a typical well [ton/year/well]

iV is the volumetric bleed rate from device i [MCF/hr/device]

Ni is the total number of device i owned by the participating companies tannual is the number of hours per year that devices were operating [8760 hr/yr] P is the atmospheric pressure [1 atm] R is the universal gas constant [0.082 L-atm/mol-K] MWgas is the molecular weight of the gas [g/mol] T is the atmospheric temperature [298 K] fj is the mass fraction of pollutant j in the vented gas


Formation-wide pneumatic device emissions were estimated according to Equation 14:

welljpneumaticTOTALpneumatic NEE ,, Equation (14)

where: Epneumatic,TOTAL is the total pneumatic device emissions of pollutant j in the formation [ton/yr] Epneumatic,j is the pneumatic device emissions of pollutant j for a single typical well [ton/yr/well] Nwell is the total number of active wells in the formation for a given calendar year

Input Data

In the absence of Haynesville Shale specific data, pneumatic devices data were taken from the East Texas Basin data as identified in Bar-Ilan, et al (2008b). Table 4-14 shows the data used to estimate pneumatic devices emissions.

January 2013

47

Table 4-14. Pneumatic devices emissions estimation inputs.

Device Type No. of

Devices* Bleed Rate (SCF/hr)*

Liquid level controller 2 31

Positioner - 15

Pressure controller 1 17

Transducer - 14

Other - -

Totals 3 * source: Bar-Ilan et al. 2008b - East Texas Basin

4.4.8 Heaters

Methodology

Heaters and boilers in use at natural gas production facilities are generally natural gas-fired external combustors. They are typically used as either separator heaters (to provide heat input to the separators), or as tank heaters (to maintain tank temperatures). It should be noted that this source category considers only tank and separator heaters, not heaters or boilers used in dehydrators. This latter usage is covered under the dehydrator source category description below. It should also be noted that this category does not consider heaters in use at large central facilities such as gas processing plants.

Heaters are primarily considered a NOx emissions source category, although they are also a minor source of CO and VOC emissions. Heater emissions are calculated on the basis of the emissions factor of the heater, and the annual flow rate of gas to the heater. The annual gas flow rate is calculated from the BTU rating of the heater and the local BTU content of the gas.

The basic methodology for estimating emissions for a single heater is shown in Equation 15:

2000106

local

annualheaterheaterheater

HV

hctQEFE

Equation (15)

where: Eheater is the emissions from a given heater [ton/yr] EFheater is the emission factor for a heater for a given pollutant [lb/million scf] Qheater is the heater MMBTU/hr rating [MMBTUrated/hr] HVlocal is the local natural gas heating value [MMBTUlocal/scf] tannual is the annual hours of operation [hr/yr] hc is a heater cycling fraction to account for the fraction of operating hours that the heater is firing (if available)


Formation-wide heater emissions are estimated by determining the typical number of heaters per well and scaling up by well count. This is shown in Equation 16:

January 2013

48

2000, TOTALheaterheaterTOTALheater WNEE

Equation (16)

where: Eheater,TOTAL is the total heater emissions in the formation [ton/yr] Eheater is the total emissions from a single heater [lb/yr] WTOTAL is the total number of wells in the formation Nheater is the typical number of heaters per well in the formation

Input Data

Heater data were taken from operator survey responses and the East Texas Basin data as identified in Bar-Ilan, et al (2008b). Table 4-15 shows the data used to estimate heater emissions.

Table 4-15. Heater emissions estimation inputs. Property Value Source

Type separator/tank Bar-Ilan et al. 2008b - East Texas Basin

Typical Fuel Natural Gas Bar-Ilan et al. 2008b - East Texas Basin

No. per Typical Well Setup 0.82 Survey

Heater MMBtu Rating (MMBTU/hr) 0.46 Survey

Annual Activity (hours) 2,200 Survey

Heater Cycling 1.00 Bar-Ilan et al. 2008b - East Texas Basin

Emission Factors

(lb/MMscf)

NOx 100 Bar-Ilan et al. 2008b - East Texas Basin

VOC 5.50 Bar-Ilan et al. 2008b - East Texas Basin

CO 84 Bar-Ilan et al. 2008b - East Texas Basin

4.4.9 Dehydrators

Methodology

Dehydrators are devices used to remove excess water from produced natural gas prior to transmission into a pipeline or to a gas processing facility. These wellhead devices are normally only used in regions where there are significant concentrations of water in the gas that cannot be removed by separators. Thus their usage is highly localized depending on the composition of the gas. There are both liquid desiccant and solid desiccant dehydrators, but in practice liquid desiccant dehydrators are overwhelmingly used. The liquid desiccant is typically either glycol, diethylene glycol (DEG) or triethylene glycol (TEG). Glycol dehydrators have two emissions sources: the still vent from which some fugitive gas is emitted; and the reboiler which is essentially a heater and has similar emissions characteristics to a heater. For both still vent and heater emissions from dehydrators, emissions factors are typically developed using the process simulation software GLYCalc, developed by the Gas Research Institute. Due to lack of Haynesville specific per-dehydrator or per-unit production emissions factors for dehydrator still vents and reboilers, regional emissions factors per unit production from Bar-Ilan et al. 2008b for the East Texas Basin were used in this analysis. These per-unit production emissions factors are used to directly estimate regional dehydrator emissions.

January 2013

49

The basic methodology for estimating formation-wide emissions from dehydrator still vents is shown in Equation 17:

6102

formationdehydrator

dehydrator

PEFE

Equation (17)

where: Edehydrator is the formation-wide emissions from dehydrators [ton-VOC/year] EFdehydrator is the emission factor per unit production [lb/MMSCF] Pformation is the formation-wide gas production [MCF/year]

Input Data

In the absence of Haynesville Shale specific data, dehydrators per unit of production emissions data were taken from the East Texas Basin data as identified in Bar-Ilan, et al. 2008b. Table 4-16 shows the data used to estimate dehydrator emissions.

Table 4-16. Dehydrator emissions estimation inputs. Property Value Source

Emission Factors

(lb/MMSCF produced gas)

NOx 0.05 Bar-Ilan et al. 2008b - East Texas Basin

CO 0.11 Bar-Ilan et al. 2008b - East Texas Basin

VOC 2.62 Bar-Ilan et al. 2008b - East Texas Basin

4.4.10 Flaring

Methodology

Flaring is used for a number of well site processes to control VOC and other emissions. Based on emissions estimation guidance for the East Texas Basin in Bar-Ilan, et al (2008b) and inputs for condensate tanks from ERG (2012), two flaring process considered for this analysis are gas dehydration and condensate tanks. Dehydrator still vent and condensate tank emissions may be controlled by flaring. The vented gas is routed to a combustor which then burns the gas to remove upwards of 95% of VOC emissions. However, these flares are themselves a source of NOx and CO emissions.

The methodology for estimating formation-wide emissions from dehydration processes and condensate tank is described in Equation 18 and Equation 19 respectively:

2000101000 6

,,

,

gasregiondehydratorflarei

ndehydratioflare

PHVQEFE

Equation (18)

where: Eflare,dehydration is the basin-wide flaring emissions from flaring of dehydrator vent gas [ton/yr] EFi is the emissions factor for pollutant i [lb/MMBtu]

January 2013

50

Qflare,dehydrator is the volume of dehydrator still vent gas flared per unit of gas produced in the basin [MCF/million MCF produced] HV is the local heating value of the gas [BTU/scf] Pregion,gas is the region-wide gas production [MCF]

2000101000 6

,tan,

tan,

condensateregionkflarei

kcondensateflare

PHVQEFE

Equation (19)

where:

Eflare,condensate is the basin-wide flaring emissions from flaring of condensate tank [ton/yr] EFi is the emissions factor for pollutant i [lb/MMBtu] Qflare,dehydrator is the volume of gas flared per unit of condensate in the region [MCF/bbl] HV is the local heating value of the gas [BTU/scf] Pbasin,ondensate is the region-wide condensate production [bbl]

Input Data

In the absence of Haynesville Shale-specific data, flaring data were taken from the East Texas Basin data as identified in Bar-Ilan, et al, 2008b and ERG (2012) study. Table 4-17 shows the data used to estimate flaring emissions.

Table 4-17. Flaring emissions estimation inputs. Property Value Source

Flared Process Dehydrator Bar-Ilan et al. 2008b - East Texas Basin

Gas Flared per Unit of Activity Surrogate (MCF flared / million MCF produced) 8.84 Bar-Ilan et al. 2008b - East Texas Basin

Emission Factors (lb/MMbtu)



Flared Process Condensate Tank ERG,2012 - Condensate Tank Oil and Gas Activity

Gas Flared per Unit of Activity Surrogate (MCF flared/bbl) 8.52

ERG,2012 - Condensate Tank Oil and Gas Activity

Emission Factors (lb/MMbtu)



4.4.11 Wellhead Compressors

Methodology

Wellhead compressor engines can represent a significant NOx area emissions source category. These engines are used to boost produced gas pressure from downhole pressure to the required pressure for delivery to a transmission pipeline. Generally these engines are natural-gas powered, using the produced gas (after some separation and dehydration) as fuel for a spark-ignited internal combustion engine. They generally operate 8760 hours per year with a

January 2013

51

minimum of down-time. Any down-time is typically associated with repairs or routine maintenance, but gas production companies attempt to minimize this down-time to the extent possible.

Wellhead compressor engines may be uncontrolled or controlled, and two distinct types are used: “rich-burn” engines that are characterized by NOx emissions factors in the range of approximately 10-20 g/bhp-hr; and “lean-burn” engines that are characterized by NOx emissions factors in the range of approximately 1.0-5.0 g/bhp-hr. The exact NOx emissions factors depend on the horsepower of the engine, the make and model, the model year of the engine, and whether the engine has been converted from a rich-burn to a lean-burn engine. Survey data was gathered on the fraction of wellhead compressor engines using exhaust emissions controls (primarily non-selective catalytic reduction – NSCR). This information was incorporated directly into the emissions estimates for wellhead compressors.

The basic methodology for estimating emissions from wellhead compressor engines is shown in Equation (20):

185,907

annuali

engine

tLFHPEFE

Equation (20)

where: Eengine are emissions from a rich-burn or lean-burn compressor engine [ton/year/engine] EFi is the emissions factor of pollutant i [g/hp-hr] (note that this may be different for NOx emissions from rich-burn vs. lean-burn engines) HP is the horsepower of the engine [hp] LF is the load factor of the engine tannual is the annual number of hours the engine is used [hr/yr]


The emissions are scaled to the formation level using the ratios of rich-burn engines to total engines, lean-burn engines to total engines, the fraction of wells with wellhead compressor engines, and the total well count in the formation, according to Equation (21):

wellheadTOTALLeanengineLeanRichengineRichTOTALengine fWECECE ,,, Equation (21)

where: Eengine,TOTAL is the total emissions from compressor engines in the formation [ton/yr] Eengine,Rich is the total emissions from a single representative rich-burn compressor engine per Equation (20) [ton/yr] Eengine,Lean is the total emissions from a single representative lean-burn compressor engine per Equation (20) [ton/yr] CRich is the fraction of wellhead compressors in the formation that are rich-burn CLean is the fraction of wellhead compressors in the formation that are lean-burn

January 2013

52

WTOTAL is the total well count in the formation fwellhead is the fraction of all wells in the formation with wellhead compressor engines

Input Data

Wellhead compressor data were taken from the East Texas Basin data as identified in Bar-Ilan, et al. (2008b), operator survey responses, and the Barnett Shale Phase I emission inventory (TCEQ, 2011a). The weighted average horsepower and fraction of rich and lean burn engines were calculated using population data from the Barnett Shale Phase I emission inventory (TCEQ 2011a). Table 4-18 shows the data used to estimate wellhead compressor emissions.

Table 4-18. Wellhead compressor emissions estimation inputs. Description Value Source

Typical Engine Operational Characteristics

Fraction of wells serviced by wellhead compressors

0.08% Survey

Fuel Type Natural Gas Bar-Ilan et al. 2008b - East Texas Basin

Rated Horsepower 127 Barnett Shale Phase I emission inventory (TCEQ, 2011a)

Annual Activity 8760 Bar-Ilan et al. 2008b - East Texas Basin

Load Factor 66% Bar-Ilan et al. 2008b - East Texas Basin

Fraction Rich Burn 95% Barnett Shale Phase I emission inventory (TCEQ, 2011a)

Fraction Lean Burn 5% Barnett Shale Phase I emission inventory (TCEQ, 2011a)

Typical Emission Factors for Rich Burn (g/bhp-hr)

NOx 2.00 EPA, 2008 - New Source Performance Standards (NSPS)

Typical Emission Factors for Rich Burn (g/bhp-hr) Typical Emission Factors for Lean Burn (g/bhp-hr)

VOC 1.00 EPA, 2008 – NSPS

CO 4.00 EPA, 2008 – NSPS

NOx 2.00 EPA, 2008 – NSPS

Typical Emission Factors for Lean Burn (g/bhp-hr)

VOC 1.00 EPA, 2008 – NSPS

CO 4.00 EPA, 2008 – NSPS

4.4.12 Condensate Tanks

Methodology

Condensate tanks can represent a significant VOC area emissions source category. The Haynesville Shale formation produces relatively small quantities of condensate.

The basic methodology for estimating emissions from condensate is shown in equation 22:

http://www.tceq.texas.gov/airquality/point-source-ei/psei.html#barnett

January 2013

53

2000

tantantan

kscondensatekscondensatekscondensate

EFPE

Equation (22)

where: Econdensate,tanks is the region-wide emissions from condensate tanks [tons/yr] EFcondensate,tank is the VOC emissions factor for condensate tanks [lb-VOC/bbl] Pcondensatetanks is the condensate production from gas wells throughput [bbl]

Input Data

Condensate tank emissions factors and other data were taken directly from a recent activity report conducted for the TCEQ (ERG, 2012). As mentioned above, producers did not provide any information that would allow for estimation of representative emission factors for condensate tank flashing losses associated with Haynesville Shale condensate production. Due to insufficient information for estimating emissions from working and breathing losses from condensate tanks, only flashing emissions were estimated. It should be noted that working and breathing loss emissions are usually significantly smaller in magnitude than flashing emissions from condensate storage tanks.

Table 4-19 shows the data used to estimate condensate tank emissions.

Table 4-19. Condensate tanks emissions input. Property Value Source

Uncontrolled Flashing Emission factor (lb/bbl) 4.22 ERG,2012 - Condensate Tank Oil and Gas Activity

Fraction of Production Controlled with Flare 10% ERG,2012 - Condensate Tank Oil and Gas Activity

Flaring Control Efficiency (%) 98% ERG,2012 - Condensate Tank Oil and Gas Activity

4.4.13 Midstream/Gas Processing Sources

Methodology

The midstream emissions are obtained from Texas and Louisiana permitted source data as explained in section 4.3 above. The base year emissions per unit of gas production from each point source are estimated according to Equation 23

2010,

2010,

region

i

iP

EE

Equation (23)

where: Ei are emissions from a midstream source due to Haynesville Shale formation production [ton/mcf]

January 2013

54

Ei,2010 are emissions from a midstream source in 2010 [ton/year] Pregion,2010 is the production from the Haynesville Shale region in 2010 [MCF/year]


Formation-wide midstream compressor engines and natural gas plant emissions are estimated by scaling up according to Haynesville Shale formation-wide gas production as shown in Equation 24:

eHaynesvilliiregion GEE X , Equation (24)

where: Eregion,i are emissions from a midstream source due to Haynesville Shale formation production [ton/year] Ei, are emissions from a midstream source per MCF [tons/MCF] GHaynesville is the Haynesville formation-wide gas production [MCF]

4.5 Forecasting Scenarios

Projections of exploration activity and gas production in the Haynesville Shale were developed for each year for the period 2012-2020. These projections form the basis for scaling the base year emissions inventory to all future years using one of the following activity factors as a projection surrogate: annual spuds; annual active well count; annual gas production; annual condensate production. The projections were developed for three scenarios: (1) a moderate (medium) growth scenario; (2) a low growth scenario; and (3) an aggressive (high) development scenario. These scenarios were developed to cover a range of possible future levels of gas exploration and production activity in the Haynesville Shale.

Many factors could affect future levels of exploration and production. Chief among these is the price of natural gas which is, in turn, affected by the state of the global economy as well as the natural gas supply from other formations in the U.S. Another important variable is the productivity of wells within the Haynesville Shale. This is measured for an individual well using a well decline curve, which tracks the amount of natural gas produced by the well as a function of time. Typically, a well will have its maximum production immediately after drilling and then productivity decreases with time as the gas reservoir is drained. Well decline curves for individual wells can be used to estimate the production of the field as a whole, since the number of producing wells in the field and the age of each well is known.

The projection scenarios were constructed for each future year using a series of steps:

1. The number of new wells drilled (spuds) were projected for each year;

2. Production estimates for each new active well were derived from well decline curves;

3. Formation-wide spuds, well counts, gas production, and condensate production were

estimated using the per-well production and considering a number of assumptions for each

scenario;

January 2013

55

4. Using the well and production projections, emissions in the base year were scaled to each

future year by assigning production surrogates to each source category (i.e. “uncontrolled

emissions”);

5. Controls associated with “on-the-books” regulations were applied to projected emissions

for each forecast year to obtain final emissions in each scenario.

The methodology for generating the activity projection scenarios is described in detail below. Each scenario was constructed considering a number of assumptions, and it should be noted that these are “best-guess” projections, and some degree of uncertainty is associated with these projections.

4.5.1 Rig Count Estimates

The number of new wells that will be drilled in the Haynesville Shale in the period 2012-2020 will depend on a broad range of economic factors, some affecting the natural gas industry as a whole and some specific to the Haynesville Shale, but all of which ultimately influence the profitability of development of the formation. The economic factors include: natural gas prices; natural gas demand; natural gas supply from other sources; productivity of Haynesville Shale wells compared with other natural gas formations; natural gas transportation infrastructure; cost of drilling in the Haynesville Shale as compared to other formations; drill rig availability; land leasing costs; and other factors.

Low Scenario

Rig counts for the low scenario were estimated using available data on drill rigs in use for 2011 and 2012 from Baker Hughes (2012). For years beyond 2012, rig counts were estimated by extrapolating recently declining rig counts into the future. A power curve was fit to weekly 2012 rig count estimates from the Baker Hughes data for January 2012 to September 2012 Table 4-20 lists the rig count estimates in each year for the low scenario. This scenario is designed as a lower bound reflecting a continued lack of financial incentive to develop new production capacity in the Haynesville.

January 2013

56

Figure 4-1. Forecast of future rig count based on 2012 rig count trends.

Table 4-20. Low scenario Haynesville Shale rig count estimates.

Year

Average Annual Rig

Count Data Source 2011 138 Feb-2011 to Dec-2011 historical data

2012 64 Jan-2012 to Sep-2012 historical data

2013 39 Power Curve Fit Forecast of Jan-2012 to Sep-2012 data 2014 34

2015 31

2016 28

2017 27

2018 26

2019 25

2020 24

High Scenario

For the high scenario it was assumed that Haynesville Shale again becomes economically attractive to producers in the near future leading to increasing activity. This scenario could be achieved in a variety of ways, such as the development of an extensive liquefied natural gas (LNG) export market or gas-to-liquids facilities such as that planned by Sasol for Lake Charles, LA (Chemical and Engineering News, 2012). This plant will use shale gas to make fuels and chemicals such as naptha, diesel, and jet fuel. 2011 rig counts estimated from Baker Hughes data as described in the first scenario were assumed to increase monotonically to a maximum of 200 in 2014. The maximum number of 200 drill rigs operating in the Haynesville Shale is

January 2013

57

consistent with the previous Haynesville Shale emission inventory (ENVIRON, 2009). The assumed growth rate in rig counts from 2011 to 2014 is the same as the historic Barnett Shale growth rate as relied upon in the previous Haynesville Shale emission inventory (ENVIRON, 2009). This scenario is designed as an upper bound that is feasible but considered unlikely in light of the current low price for natural gas.

Moderate Scenario

The moderate scenario falls between the high and low scenarios and relies upon assumptions about the market price of natural gas and the price point at which developing the Haynesville is economically attractive. The moderate scenario assumes that the current downturn in drilling in the Haynesville Shale will continue until gas prices rise to $6/MMBTU, which was approximately the price of natural gas at the time of the initial increase in activity in the Haynesville Shale. The Energy Information Administration’s (EIA) Annual Energy Outlook (AEO) predicts $6/MMBTU in 2022, ten years from 2012. For the first half of this ten year period (2012-2016), we assume that annual average drill rig counts follow the same curve as the low scenario. In 2017, we assume that drilling activity in the Haynesville Shale begins to increase at half of the historic Barnett Shale growth rate as presented in the previous Haynesville Shale Emission Inventory Report (ENVIRON, 2009). This scenario is designed as a plausible future considering both the current low price for natural gas and a recovering economy leading to growing demand for gas.

Scenario Results

Average annual rig count estimates for all scenarios are shown in Figure 4-2. These estimates are the basis of the future emission scenarios.

January 2013

58

Figure 4-2. Total annual drill rig count in the Haynesville Shale predicted for the three development scenarios in the period 2009-2020.

4.5.2 Drilling and Well Count Estimates

The number of wells drilled per year by a rig operating in the Haynesville Shale was estimated to be 15.6 based on the average operator provided estimate of 22.3 days of drill rig operation per well, assuming one day for rig moving between drilling events. Forecasted well counts were estimated based on a 2008 success factor of 55% increasing linearly to 100% in 2018 similar to ENVIRON (2009). Figure 4-3 shows the annual number of spuds while Figure 4-4 shows active well counts for 2011 to 2020 for each alternative.

January 2013

59

Figure 4-3. Haynesville shale annual spud counts by year for each scenario.

Figure 4-4. Haynesville shale active well count estimates by year for each scenario.

January 2013

60

4.5.3 Production Estimates

Using the well development estimates for each of the three scenarios and estimates for the typical gas production of a well over its lifetime, total formation-wide gas production can be calculated for the three development scenarios. This analysis requires deriving estimates of typical well production over the time period 2010-2020, during which a well’s production is expected to decline from an initial production peak.

To estimate well decline over the first 20 years of production, the Haynesville Shale wells scout reports were obtained from the LDNR website (LDNR, 2012c). 40 wells had a first production date of 2008 and these wells were selected for analysis. Detailed monthly production data was obtained from LDNR SONRIS database (LDNR, 2012d) for each well. Figure 4-5 shows an example of monthly gas production estimates for well 236242. It was assumed that the first three months of production for well 236242 were during well testing or stimulation activities and not representative of typical well decline. The first three months of production were therefore not included in the well decline analysis. Each well’s period of recorded data was evaluated and initial data inconsistent with typical well decline was removed from consideration in the well decline analysis.

Figure 4-5. Monthly gas production for well 236242 (LDNR, 2012d).

For each well, estimates of average percentage decline from the first half year of gas production in half year increments were estimated and a power curve was fit to the average percentage decline estimate across all wells (see Figure 4-6). The power curve regression was combined with the average estimate of production in the first six months of well operation

January 2013

61

(1.201 BCF) to arrive at a well decline curve. Decline estimates were strongly affected by wells with the longest period of record (24-30 months). These wells showed significantly lower production trends relative to the other wells; therefore these wells were conservatively removed from the analysis. The resulting well decline curve estimate is shown in Figure 4-7. This well decline method resulted in an estimate of 6.9 BCF of gas produced over the first 20 years of production which is in the range of other available estimates of Haynesville Shale well production. For example Chesapeake Energy (2012) estimated recovery over the life of a Haynesville Shale well at 6.5 BCF.

Figure 4-6. Average percentage decline from the first half year of gas production.

ENVIRON evaluated two alternative methods of estimating well decline; however these methods were not used to estimate gas production in this study because production estimates were high relative to published values.

1. Similar to the method used in the ENVIRON (2009) study, power curve regressions were

estimated for each well. Power curve estimates of monthly production for each well were

averaged across all wells to estimate an average decline curve. This method resulted in

estimates of cumulative production over twenty years of 10.8 BCF.

2. Power curve fitting to cumulative production estimates was also evaluated. This method

resulted in estimates of cumulative production over twenty years of 9.9 BCF.

January 2013

62

Figure 4-7. Haynesville Shale well decline estimate.

95% confidence intervals were evaluated to estimate the level of variability associated with the estimated decline curve (see Figure 4-8). The confidence intervals incorporate the variability in (1) the estimated power curve regression and (2) the estimate of gas production over the first six months of production activity. The 95% confidence intervals show production estimates with a range of 2.5 BCF to 14.8 BCF of cumulative gas production over the first 20 years of production. A wide range of well decline estimates is expected considering (1) that the period of available production records is limited to three to four years while the expected life of a Haynesville Shale well is on the order of twenty years and (2) that there are a relatively small number of wells that have been producing since 2008 which are available to include in this analysis.

January 2013

63

Figure 4-8. Haynesville Shale well decline estimate with confidence intervals.

Haynesville Shale formation-wide gas production estimates for the period 2011-2020 were obtained by multiplying the number of active wells by the appropriate annual gas production rate determined from the decline curve and the year that each well was brought on-line, and summing over all active wells. Annual gas production for each scenario is shown in Figure 4-9. Cumulative gas production is shown in Figure 4-10.

January 2013

64

Figure 4-9. Total Haynesville Shale annual gas production estimates.

Figure 4-10. Total Haynesville Shale cumulative gas production estimates.

January 2013

65

The 2020 cumulative production figures in Figure 4-10 show a large difference between the low (26.3 TCF), moderate (30.7 TCF) and high (96.8 TCF) scenarios. The low and moderate scenario cumulative production estimates are well below published estimates of the technically recoverable reserves in the Haynesville Shale while the aggressive scenario gas production is higher than some estimates of technically recoverable reserves. For example, the EIA (2011c) estimated 75 TCF in recoverable gas. Note that the timeline in Figure 4-11 only extends to 2020, and the Haynesville Shale would presumably still be under development after 2020.

Relatively small amounts of condensate are produced by Haynesville Shale wells. Condensate production in 2011 was forecasted to future years by assuming the same ratio for forecast year to 2011 base year production as for gas production. Figure 4-11 presents annual estimates of condensate production from the Haynesville Shale formation.

Figure 4-11. Total Haynesville Shale annual condensate production estimates.

4.6 Future Year Control Methodology

This methodology considered any “on-the-books” federal or state regulations that would affect the emissions projections.

Table 4-21 below lists the “on-the-books” federal and state regulations that affect emissions source categories in the oil and gas industry, and the action taken to adjust the 2011 emissions inventory appropriately. A more detailed description follows including the methodology used to address each of these regulations as they affected the uncontrolled future year Haynesville Shale formation emissions projections.

January 2013

66

Table 4-21. Summary of federal and state “on-the-books” regulations affecting the oil and gas source categories considered in this inventory.

Source Category Regulation Enforcing Agency

Effective Date

Implementation in the Haynesville Shale formation Emissions Projections

Federal

Drill Rigs, Fracing Nonroad engine Tier standards (1-4) (EPA, 2008)

US EPA Phase in from 1996 - 2014

EPA NONROAD model used to create county-level control factors for the drill rig SCC to account for fleet turnover.

All New Spark-Ignited Stationary Engines

New Source Performance Standards (NSPS) (EPA, 2008)

US EPA Phase in from 2008 - 2011

Control factors developed for wellhead engines considering the specific composition of engines in the inventory.

Pneumatic Controllers

Subparts OOOO: New Source Performance Standards

US EPA October 15, 2013

Control factors developed for pneumatic controllers considering the low-bleed deice rate.

State

Engines East Texas Combustion Rule

TCEQ Effective from 2010

Control factors developed for wellhead engines considering the specific composition of engines in the inventory.

4.6.1 Nonroad Diesel Engine Standards and Fuel Sulfur Standards

The EPA NONROAD2008 model was run with diesel fuel inputs of 32 ppm for 2009 to 2013, 20 ppm for 2014 and 15 ppm for 2015 to 2020 (EPA,2009). The model outputs were used to develop emissions per unit population for “other oil field equipment”, “Grader”, “Tractors/Loaders/Backhoes” and “Crawler Tractors” (SCC 2270010010, 2270002048,

2270002066, 2270002069) for the calendar years 2011 to 2020. These emissions per unit population reflect the predicted fleet mix of engines – for various tier standards from baseline uncontrolled engines through Tier IV engines – and are used as a representation of fleet turnover for drilling rigs, fracing engines and well pad construction equipment. The ratios of per unit emissions in a future year to those in 2011 for each were taken to be the control factors accounting for federal non-road tier standards.

4.6.2 Spark Ignition Engines

New Source Performance Standards for Stationary Spark-Ignited Engines

The EPA promulgated performance standards covering new stationary, spark-ignited engines of various horsepower classes. The regulation is assumed to apply to central compressor engines, wellhead and lateral compressor engines, and other engines that are stationary, spark-ignited natural gas engines. The regulation requires new engines of various horsepower classes to meet increasingly stringent NOx and VOC emission standards over the phase-in period of the regulation.

For engines less than 25 horsepower, Table 4-22 shows the requirements of the NSPS regulation.

January 2013

67

Table 4-22. Federal NSPS emissions standards for engines less than 25 horsepower.

HP Rangea

Emissions Standards Requirement in (g/hp-hr)b

HC + NOx NMHC + NOxc CO

≤ 25 Hp

Class I 16.1 (12.0) 14.8 (11.0) 610 (455)

Class I –A 50-37 - -

Class I –B 40 (30) 37 (27.6)

Class II 12.1 (9.0) 11.3 (8.4) a.

Class I-A: Engines with displacement less than 66 cubic centimeters (cc); Class 1-B: Engines with displacement greater than or equal to 66cc and less than 100cc; Class I: Engines with displacement greater than or equal to 100 cc and less than 225 cc b.

Modified and reconstructed engines manufactured prior to July 1, 2008, must meet the standards applicable to engines manufactured after July 1, 2008 c.

NMHC+NOX standards are applicable only to natural gas fueled engines at the option of the manufacturer, in lieu of HC+NOX standards

For engines in the horsepower range 25 – 100 horsepower, Table 4-23 shows the requirements of the NSPS regulation.

Table 4-23. Federal NSPS emissions standards for engines greater than 25 horsepower but less than 100 horsepower.

HP Range Manufacture Date

Emissions Standards Requirement (g/hp-hr)

HC + NOx CO 25<HP<100 1-Jul-08 3.8 6.5

1-Jul-08 (severe duty) 3.8 200

For engines in the horsepower range 100 – 1,350 horsepower, Table 4-24 shows the requirements of the NSPS regulation.

Table 4-24. Federal NSPS emissions standards for engines greater than 25 horsepower but less than 100 horsepower.

Engine Type and Fuel HP Range Manufacture

Date

Emissions Standards Requirement (g/hp-hr)

NOx CO VOC

Non-Emergency SI Natural Gas and Non-Emergency SI Lean Burn LPG

100≤HP<500 1-Jul-08 2 4 1

1-Jan-11 1 2 1

Non-Emergency SI Lean Burn Natural Gas and LPG

500≥HP<1350 1-Jan-08 2 4 1

1-Jul-10 1 2 1

Non-Emergency SI Natural Gas and Non-Emergency SI Lean Burn LPG (except lean burn 500≥HP<1350)

HP≥500 1-Jul-07 2 4 1

January 2013

68

East Texas Combustion Rule

TCEQ promulgated rule §117.3310 in June 2007. Section 117.3310(a) specifies the NOx emission specifications for stationary gas-fired reciprocating internal combustion engines. Beginning in 2010, stationary gas-fired rich-burn engines with a rated capacity between 240 (inclusive) and 500 horsepower in all Texas Counties within the Haynesville Shale area are subject to the East Texas Combustion Rule. The East Texas Combustion Rule limits NOx emissions from these engines to 1.00 g/hp-hr.

Implementation

The East Texas Combustion Rule does not require engines with a maximum rated horsepower capacity of less than 240 horsepower to meet the 1 g/bhp-hr NOx emission standard. The average horsepower for wellhead compressor engines used in this inventory is 127 horsepower and hence is exempt from the East Texas Combustion Rule. However, the effect of NSPS on wellhead compressor engines is estimated for the added natural gas compressors.

Midstream source emissions include spark-ignited engines used at compressor stations and gas processing facilities which also would be expected to be affected by NSPS and East Texas Combustion Rule requirements, however, given the lack of engine specific emissions, horsepower rating, and activity data for midstream sources, estimating expected decreases to emissions as a result of this requirement would be highly speculative and thus has not been performed. At this time, midstream emissions which were not controlled from the base year can be considered to be conservative.

4.6.3 Subparts OOOO: New Source Performance Standards regulating VOC, SO2 and new affected facilities

In April 2012, EPA promulgated the Subpart OOOO rules which regulate VOC emissions from equipment leaks and SO2 emissions from sweetening units at onshore gas processing plants. New oil and natural gas facilities are affected by this rule, including well completions at new hydraulically fractured natural gas wells and at existing wells that are fractured or refractured. The Subpart OOOO standards, shown in Table 4-25 and Table 4-26, apply to affected facilities that begin construction, reconstruction or modification after August 23, 2011.

Federal regulations were previously focused mainly on large sources such as gas processing plants and some internal combustion engines. Some engines used at the well-site such as compressors and artificial lift engines have been subject to emission limits for NOx, particularly for new or modified sources of a wide range of sizes. With the introduction of Subpart OOOO requirements, new and modified small sources at the well-site that were previously unregulated are now subject to mandatory controls. These emissions requirements affect a broader range of source categories, including well completions, pneumatic devices, condensate tanks and dehydrators, and provide guidance on mitigation measures to meet those standards. Such is the case for well completion venting from wells that have been hydraulically fractured, a significant source of VOC emissions, which must now be controlled at the least by a flaring device, prior to January, 1st 2015, and must apply reduced emissions completions (RECs) to capture the majority of the gas from January, 1st 2015 onward. This particular requirement will

January 2013

69

significantly reduce emissions associated with unconventional gas formations such as shale and tight sands where completions and recompletions (well stimulation that may be necessary after the first year of production) typically involve hydraulic fracturing. Subpart OOOO also affects emissions sources at gas processing plants previously unregulated by federal standards such as compressor leaks, gas sweeting units and small glycol dehydrators.

Implementation

Insufficient information was available on green completions in the Haynesville Shale to evaluate potential effects on emissions of Subpart OOOO requirements. Due to the lack of available data, it was not possible to determine whether Haynesville wells specifically would meet the feasibility requirements for green completions under Subpart OOOO. Hence, a conservative assumption was made that no controls due to Subpart OOOO would be applied to well completion emissions. Similarly, insufficient information on condensate tanks was available to evaluate the applicability of Subpart OOOO controls. It was assumed that for Haynesville Shale condensate tanks. 10 % of annual condensate production (ERG, 2012) is controlled by flaring in the base and all forecasted years. The Subpart OOOO requirement for low bleed pneumatic controllers is applicable beginning October 15, 2013. It was assumed that all new wells coming online after October 15, 2013 will have low bleed pneumatic controllers with a bleed rate of 6 cfh . Subpart OOOO effects on midstream sources were not incorporated into the Haynesville Shale emission inventory due to insufficient data which would allow for the evaluation of the effect of these requirements on midstream emissions sources.

January 2013

70

Table 4-25. Subpart OOOO requirements for sources at the well site. Item No.

Industry it applies to Affected Facility/Source Pollutant Standard

1 Natural Gas Hydraulically fractured wildcat (exploratory) and delineation gas wells

VOC Route flowback emissions to completion combustion device

2 Natural Gas Hydraulically fractured low pressure gas wells, non-wildcat and non-delineation gas wells


3 Natural Gas All other hydraulically fractured and refractured gas wells (Prior to Jan 1, 2015)


4 Natural Gas All other hydraulically fractured and refractured gas wells (On and after Jan 1, 2015)

VOC Must employ Reduced Emissions Completions (REC) in combination with use of a completion combustion device to control gas not suitable for entering the flow line.

5 Oil and Natural Gas

Continuous bleed natural gas-driven pneumatic controllers with a bleed rate greater than 6 scfh

VOC Natural gas bleed rate less than 6 scfh

6 Oil and Natural Gas

Condensate storage tanks VOC Storage vessels with VOC emissions equal to or greater than 6 tpy must reduce emissions by at least 95 percent. This can be accomplished by routing emissions to a combustion device.

January 2013

71

Table 4-26. Subpart OOOO requirements for sources at natural gas processing plants.

Item No.

Industry it applies to Affected Facility/Source Pollutant Standard

1 Natural Gas Centrifugal compressors with wet seals VOC Reduce fugitive emissions by 95 percent. This can be accomplished through flaring, or by routing captured gas back to a compressor suction or fuel system. Rule does not apply to compressors using dry seals.

2 Natural Gas Reciprocating compressors VOC Change rod packing after 26,000 hours of operation or after 36 months to prevent fugitive emissions.

3 Natural Gas Continuous bleed natural gas-driven pneumatic controllers

VOC Natural gas bleed rate of zero

4 Natural Gas Condensate storage tanks VOC Storage vessels with VOC emissions equal to or greater than 6 tpy must reduce emissions by at least 95 percent. This can be accomplished by routing emissions to a combustion device.

5 Natural Gas Equipment leaks at on-shore natural gas processing plants

VOC Application of leak detection and repair (LDAR) program. LDAR describes procedures and leak thresholds established by 40 CFR 60, subpart VVa. Subpart VVa lowers the leak definition for valves from 10,000 ppm to 500 ppm.

6 Natural Gas Sweetening units at on-shore natural gas processing plants

SO2 Reduce SO2 emissions to 99.9 percent based on sulfur feed rate and sulfur content of acid gas. This requirement applies to units with a sulfur production rate of at least 5 long tons per day.

January 2013

72

5.0 RESULTS

Summary results of the Haynesville Shale regional emissions inventory are presented below for NOx, VOC and CO emissions. The results are presented in detail for calendar year 2012 only as this is the future year for which regional photochemical ozone modeling will be conducted by NETAC. However, the inventory has been generated for each calendar year in the period 2011-2020 and the summary inventory results are presented for these additional calendar years, including the base year of 2011.

5.1 Detailed Emissions Inventory for 2012

Results of the Haynesville Shale regional emissions inventory for 2012 are presented below in tabular (Table 5-1) and graphical formats. Table 5-1 indicates that 2012 NOx emissions in Northeast Texas and Northwest Louisiana due to development in the Haynesville Shale range from 32 tons/day in the low and moderate development scenarios to 50 tons/day in the high (aggressive) scenario and VOC emissions from the Haynesville range from 24 tons/day in the low and moderate scenarios to 36 tons/day in the aggressive scenario. CO emissions range from 20 tons/day to 31 tons/day. It should be noted that the moderate scenario follows the same growth as low development scenario until 2017. Hence, 2012 emissions for low and moderate scenarios are identical.

Figure 5-1, which illustrates the breakdown of emissions by source category, shows that for the moderate scenario, drill rigs and midstream compressor station and natural gas plant emissions are estimated to account for 96% of NOx emissions from the Haynesville Shale. Midstream compressor station and natural gas plant contribute the largest amount of NOx, making up 80% of the total NOx emissions. Midstream compressor station and natural gas plant NOx emissions total 25 tons/day for low and moderate scenarios and 34 tons/day for the high development scenarios (Table 5-1).

As shown in Figure 5-2 for the moderate scenario, dehydrator and midstream compressor station and natural gas plant emissions are projected to account for 83% of VOC emissions while pneumatic devices and completion venting each are estimated to account for 13% of VOC emissions. Figure 5-3 shows that, for the moderate scenario, drill rigs and midstream compressor station and natural gas plant emissions are projected to account for 93% of CO emissions from the Haynesville Shale formation.

2012 emissions for the low, moderate, and aggressive (high) development scenarios as compared in Figure 5-4 shows that relative to the moderate scenario, NOx, VOC and CO emissions are identical for the low scenario and NOx emissions are 59% higher, VOC emissions are 47% higher and CO emissions are 57% higher for the high scenario. For all three scenarios, midstream compressor station and natural gas plant emissions are the largest sources of NOx emissions.

January 2013

73

Table 5-1. 2012 emissions (TPD) of NOx, VOC, and CO by scenario for the Haynesville Shale formation.

Pollutant Category Low Moderate High

NOx

Blowdowns 0.00 0.00 0.00

Completion Venting 0.00 0.00 0.00

Dehydrator 0.21 0.21 0.28

Drill Rigs 5.04 5.04 12.91

Flaring 0.00 0.00 0.00

Fracing 0.60 0.60 1.53

Fugitives 0.00 0.00 0.00

Heaters 0.39 0.39 0.52

Pneumatic Devices 0.00 0.00 0.00

Wellhead Compressors 0.01 0.01 0.01

Midstream CS and NGP 25.11 25.11 34.41

Wellpad Construction 0.16 0.16 0.42

Condensate Tank 0.00 0.00 0.00

NOx Total 31.51 31.51 50.08

VOC

Blowdowns 0.08 0.08 0.10


Dehydrator 10.36 10.36 14.20

Drill Rigs 0.29 0.29 0.73

Flaring 0.00 0.00 0.00

Fracing 0.09 0.09 0.24

Fugitives 0.09 0.09 0.12

Heaters 0.02 0.02 0.03






VOC Total 24.32 24.32 35.73

CO

Blowdowns 0.00 0.00 0.00


Dehydrator 0.41 0.41 0.57

Drill Rigs 2.84 2.84 7.29

Flaring 0.01 0.01 0.02

Fracing 0.36 0.36 0.94

Fugitives 0.00 0.00 0.00

Heaters 0.32 0.32 0.44






CO Total 19.62 19.62 30.79

January 2013

74

Figure 5-1. 2012 moderate development scenario Haynesville Shale formation NOx proportional contributions by source category.



2012 Flaring0.01%

2012 Fracing1.89%

2012 Heaters1.23%


79.70%


0.02%


0.51%

2012 Haynesville Shale NOx Emissions Contribution By Source Category for Moderate Scenario


January 2013

75

Figure 5-2. 2012 moderate development scenario Haynesville Shale formation VOC proportional contributions by source category.

2012 Blowdowns

0.31%

2012 Completion Venting6.72%

2012 Dehydrator42.60% 2012 Drill Rigs

1.17%

2012 Fracing0.39%

2012 Fugitives0.37%

2012 Heaters0.09%


40.41%

2012 Pneumatic

Devices6.22%


0.02%


0.09%

2012 Condensate

Tank1.61%

2012 Haynseville Shale VOC Emissions Contribution By Source Category for Moderate Scenario


January 2013

76

Figure 5-3. 2012 moderate scenario Haynesville Shale formation CO proportional contributions by source category.



2012 Flaring0.07%

2012 Fracing1.86%

2012 Heaters1.66%


79.38%


0.08% 2012 Wellpad Construction

0.35%

2012 Haynesville Shale CO Emissions Contribution By Source Category for Moderate Scenario


January 2013

77

Figure 5-4. 2012 Haynesville Shale formation emissions of NOx, VOC, and CO by scenario and source category.

5.2 Emissions Inventories for Calendar Years 2011-2020

Results of the Haynesville Shale regional emissions inventory for the period 2011-2020 are presented below as a series of bar graphs for the moderate scenario. A quantitative emissions summary is presented in Table 5-2 with results for all scenarios and percent changes from the moderate scenario for the low and high (aggressive) scenarios. When reviewing these results, changes in emissions from year to year by source category can be interpreted based on the understanding that the change in emissions from 2011 to future years for each category is a function of two parameters for any given category, (1) the change in the surrogate associated with the category, and (2) emissions controls (if any) associated with the category.

Figure 5-5 shows that moderate scenario NOx emissions are projected to decrease by 30% for 2016 relative to 2011 and increase by 30% for 2020 relative to 2011. By 2020, development in the Haynesville Shale results in more than 46 tons/day of NOx emitted in northeast Texas and northwest Louisiana. Notably, while the number of spuds decreases by 43%, drill rig emissions decrease by an even larger amount (73%) because of turnover in the drill rig engine fleet that results in replacement of older engines with higher Tier, cleaner-burning engines. Midstream compressor station and natural gas plant NOx emissions account for most of the increase in

0

10

20

30

40

50

60

Low Moderate High Low Moderate High Low Moderate High

CO NOx VOC

Em

issio

ns

(T

PD

)2012 Haynesville Shale NOx, VOC, and CO Emissions by scenario and source

category

Condensate Tank

Wellpad Construction

Midstream CS and NGP


Pneumatic Devices

Heaters

Fugitives

Fracing

Flaring

Drill Rigs

Dehydrator

Completion Venting

Blowdowns

January 2013

78

overall NOx emissions. Figure 5-6 shows that moderate scenario VOC emissions are projected to increase by 66% from 2011 to 2020. VOC emission increases are primarily due to increases in dehydrator and midstream compressor station and natural gas plant VOC emissions, though pneumatic devices and condensate tanks among other categories also contribute to VOC emission increases. Figure 5-7 shows that moderate scenario CO emissions are projected to increase by 31% from 2011 to 2020. Midstream compressor station and natural gas plants account for most of the increases in CO emissions.

Table 5-2. 2011 to 2020 emissions of NOx, VOC, and CO by scenario for the Haynesville Shale region and percent difference from the moderate scenario.

Pollutant Year

Emissions (TPD) Percent Difference from

Moderate

Low Moderate High Low High NOx 2011 35.51 35.51 35.51 0% 0%

2012 31.51 31.51 50.08 0% 59%

2013 28.59 28.59 69.96 0% 145%

2014 26.35 26.35 86.24 0% 227%

2015 25.38 25.38 99.39 0% 292%

2016 24.92 24.92 110.85 0% 345%

2017 24.74 27.33 121.61 -9% 345%

2018 24.78 32.48 132.05 -24% 307%

2019 24.83 38.94 141.16 -36% 262%

2020 24.80 46.17 148.65 -46% 222%

VOC 2011 23.26 23.26 23.26 0% 0%

2012 24.32 24.32 35.73 0% 47%

2013 23.39 23.39 52.78 0% 126%

2014 21.98 21.98 66.87 0% 204%

2015 21.46 21.46 78.71 0% 267%

2016 21.29 21.29 89.07 0% 318%

2017 21.30 23.18 98.79 -8% 326%

2018 21.47 27.39 108.17 -22% 295%

2019 21.63 32.73 116.39 -34% 256%

2020 21.69 38.71 123.12 -44% 218%

CO 2011 21.76 21.76 21.76 0% 0%

2012 19.62 19.62 30.79 0% 57%

2013 17.90 17.90 43.09 0% 141%

2014 16.54 16.54 53.41 0% 223%

2015 15.94 15.94 61.67 0% 287%

2016 15.66 15.66 68.84 0% 340%

2017 15.54 17.06 75.55 -9% 343%

2018 15.57 20.18 82.04 -23% 307%

2019 15.59 24.09 87.67 -35% 264%

2020 15.55 28.43 92.17 -45% 224%

January 2013

79

Figure 5-5. 2011 to 2020 moderate scenario Haynesville Shale NOx emissions by source category.

0

5

10

15

20

25

30

35

40

45

50

2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

NO

x Em

issi

on

s (T

PD

)

Year

2011 to 2020 moderate scenario Haynesville Shale NOx emissions by source category



Pneumatic Devices


Heaters

Fugitives

Fracing

Flaring

Drill Rigs

Dehydrator

Condensate Tank

Completion Venting

Blowdowns

January 2013

80

Figure 5-6. 2011 to 2020 moderate scenario Haynesville Shale VOC emissions by source category.

0

5

10

15

20

25

30

35

40

45

2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

VO

C E

mis

sio

ns

(TP

D)

Year

2011 to 2020 moderate scenario Haynesville Shale VOC emissions by source category



Pneumatic Devices


Heaters

Fugitives

Fracing

Flaring

Drill Rigs

Dehydrator

Condensate Tank

Completion Venting

Blowdowns

January 2013

81

Figure 5-7. 2011 to 2020 moderate scenario Haynesville Shale CO emissions by source category.

0

5

10

15

20

25

30

2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

CO

Em

issio

ns (

TP

D)

Year

2011 to 2020 moderate scenario Haynesville Shale CO emissions by source category



Pneumatic Devices


Heaters

Fugitives

Fracing

Flaring

Drill Rigs

Dehydrator

Condensate Tank

Completion Venting

Blowdowns

January 2013

82

5.3 Comparison of Emissions

The base year of the emission inventory has been updated from 2009 in the original ENVIRON (2009) study to 2011 in this study. Updates to key emission inventory inputs including drilling time, gas composition analysis, wellhead compressor engines usage, and well decline and projection scenarios have resulted in an updated Haynesville Shale emission inventory that differs significantly from the original 2009 study estimates.

Table 5-3 shows base year emissions per well for the updated and original emission inventory. As evident from Table 5-3, dehydrator VOC emissions from the original emission inventory are increased due primarily to a correction of the dehydrator VOC emission rate and secondarily to an increase in per well gas production for the base year. Increased flaring emissions are due to the combination of increased per well gas production and the addition of condensate tank flaring emissions. Fugitive component VOC emissions decreased relative to the previous ENVIRON (2009) inventory due to an updated natural gas composition.

The Haynesville-specific data from operators suggests a reduction in drilling, heaters and wellhead compressor engines emissions. A decrease is observed in the 2010 TCEQ point source emissions per unit of produced gas for the updated emission inventory relative to the original inventory. Although activity data used for venting sources were unchanged from the original emission inventory, VOC emissions from all venting source categories were lower due to the lower average VOC content of the vented gas based on the Haynesville specific gas composition analysis relative to the VOC content of the gas composition used in the original ENVIRON (2009) study. Relative to the original ENVIRON (2009) study, the well decline curve estimates of gas production are approximately three times higher over the first 20 years of production while the point source emissions per unit gas production were lower by 79%, 64%, and 81% for NOx, VOC and CO, respectively. The combination of higher per well production and lower tons of midstream emissions per unit gas production resulted in lower per well emissions for midstream compressor stations and natural gas processing plants.

Table 5-4 provides emissions for all three scenarios for the original and updated emission inventories. The original emission inventory captures 2009 to 2020 emissions while the updated emission inventory projects 2011 to 2020 emissions. NOx and CO emissions per day for all scenarios are higher in the original emission inventory, while VOC emissions per day are lower in the original emission inventory for all scenarios.

Table 5-3. Percentage change in per well base year emissions from previous ENVIRON (2009) and the updated emission inventory.

Category

Percentage Change Absolute Change (tons/well)

NOx VOC NOx VOC Blowdowns 0.0% -30.2% 0.00 -0.004

Completion Venting 0.0% -30.2% 0.00 -0.26

Dehydrator 126.6% 5558.0% 0.01 1.26

Drill Rigs -93.7% -97.2% -29.93 -3.88

Flaring 136.5% 0.0% 0.00 0.00

Fracing 0.0% 0.0% 0.00 0.00

January 2013

83

Category

Percentage Change Absolute Change (tons/well)

NOx VOC NOx VOC Fugitives 0.0% -21.6% 0.00 -0.003

Heaters -54.4% -54.4% -0.05 0.00

Midstream CS and NGP -51.9% -17.9% -3.35 -0.27

Pneumatic Devices 0.0% -30.2% 0.00 -0.07

Wellhead Compressors -98.1% -98.0% -0.06 -0.03

Wellpad Construction No emission estimates in previous emission inventory

Condensate Tank No emission estimates in previous emission inventory

Table 5-4. Original ENVIRON (2009) and updated emission inventory presented in this report for all scenarios (tons per day).

Pollutant Year

Original ENVIRON (2009) Emission Inventory - Emissions (TPD)

Updated Emission Inventory- Emissions (TPD)

Low Moderate High Low Moderate High

NOx

2009 56.69 56.69 62.39 No Estimates

2010 60.44 67.34 88.35 No Estimates

2011 60.69 74.77 113.49 35.51 35.51 35.51

2012 60.64 81.72 139.84 31.51 31.51 50.08

2013 60.59 88.54 167.73 28.59 28.59 69.96

2014 60.49 95.14 187.12 26.35 26.35 86.24

2015 60.63 102.00 202.92 25.38 25.38 99.39

2016 61.00 109.21 217.57 24.92 24.92 110.85

2017 61.60 116.85 231.55 24.74 27.33 121.61

2018 62.43 122.31 245.13 24.78 32.48 132.05

2019 63.12 125.11 257.08 24.83 38.94 141.16

2020 63.70 127.06 267.08 24.80 46.17 148.65

VOC

2009 9.55 9.55 10.88 No Estimates

2010 10.80 11.93 16.34 No Estimates

2011 11.50 13.96 22.16 23.26 23.26 23.26

2012 12.15 16.04 28.55 24.32 24.32 35.73

2013 12.83 18.27 35.62 23.39 23.39 52.78

2014 13.61 20.78 41.70 21.98 21.98 66.87

2015 14.36 23.34 46.86 21.46 21.46 78.71

2016 15.14 26.11 51.74 21.29 21.29 89.07

2017 15.98 29.09 56.45 21.30 23.18 98.79

2018 16.85 31.73 61.05 21.47 27.39 108.17

2019 17.66 33.74 65.19 21.63 32.73 116.39

2020 18.38 35.43 68.74 21.69 38.71 123.12

CO

2009 36.61 36.61 40.96 No Estimates

2010 39.81 44.23 59.43 No Estimates

2011 40.29 49.40 77.35 21.76 21.76 21.76

2012 40.54 54.30 96.36 19.62 19.62 30.79

2013 40.52 58.75 116.13 17.90 17.90 43.09

2014 40.92 63.78 131.64 16.54 16.54 53.41

2015 41.15 68.51 143.98 15.94 15.94 61.67

2016 41.48 73.37 155.27 15.66 15.66 68.84

2017 41.91 78.43 165.92 15.54 17.06 75.55

2018 42.47 82.15 176.13 15.57 20.18 82.04

2019 42.82 83.91 184.89 15.59 24.09 87.67

January 2013

84

Pollutant Year

Original ENVIRON (2009) Emission Inventory - Emissions (TPD)

Updated Emission Inventory- Emissions (TPD)

Low Moderate High Low Moderate High

2020 42.84 84.53 191.66 15.55 28.43 92.17

A comparison is presented of the baseline emissions estimated for the Haynesville Shale presented in this report with the TCEQ special emission inventory for the Barnett Shale (TCEQ, 2010, TCEQ, 2011a). TCEQ has developed a detailed inventory of emissions for year 2009 from Barnett Shale exploration and production sources through a mandatory survey process. Since the Special Inventory for the Barnett Shale did not include drilling rig emissions, results for Barnett Shale producing counties from TCEQ’s statewide drilling rig 2008 emissions inventory (ERG, 2009) are combined with the special inventory 2009 emissions for comparison purposes. It should be noted that the TCEQ’s statewide emission inventory did not estimate well completion emissions. Table 5-5 shows Haynesville’s annual baseline emissions and TCEQ Barnett Shale emissions with gas production statistics included for comparison purposes.

Table 5-5. Summary of Haynesville base year emissions and 2009 Barnett Shale emissions with regional production statistics.

Source Data Regional Active

Well Counts

Regional Gas Production

(BCF/yr)

Emissions (TPY)

NOx VOC CO 2009 Barnett Shale 13,902

a 1,773

b 29,443

b 20,427

b 2,155

b

Haynesville Shale– 2011 Base year

2,501 2,444 12,962 8,488 7,942

a Well count as of May, 2010 per TCEQ (2010)

b per TCEQ (2011a)

The TCEQ special inventory for the Barnett Shale provides emissions by source category which is compared with the Haynesville Shale baseline emissions by source category in Table 5-5.

For the Barnett Shale inventory, NOx emissions are mainly from compressor engines and drilling while condensate tank are the major source of VOC emissions (about half of the total) in 2009, followed by fugitive leak emissions. The Haynesville Shale inventory shows midstream compressor stations and natural gas plant and drill rigs as the major NOx contributors whereas VOC emissions are mainly from dehydrators and midstream compressor stations and natural gas plants. The Barnett Shale inventory does not account for emissions from midstream sources which dominate Haynesville Shale NOx and VOC emissions. It is observed that condensate tank emissions are much more significant in the Barnett Shale compared to the Haynesville Shale due to much higher condensate production rates in the Barnett relative to the Haynesville. Wellhead compressor emissions are significantly higher in the Barnett Shale as a result of higher fractional usage of engines, and a larger resulting engine population, than in the Haynesville Shale. This finding is consistent with information provided by operators in the Haynesville Shale indicating that Haynesville wells produce at high pressure requiring very little additional wellhead compression.

January 2013

85

Table 5-6. By source category emissions for Barnett Shale and Haynesville Shale emission inventory.

Source Categories

Annual Emissions (tpy)

Barnett Shale - 2009 Haynesville Shale - 2011 Base Year

NOX VOC CO NOX VOC CO

Amine Units 0 6 0 - - -

Boilers 13 1 0 - - -

Condensate Tank 0 11,549 0 0 121 0

Compressor Engines

(wellhead and lateral) 19,013 1,236 0 3 1 6

Midstream Compressor Station and Gas Plants - - - 7,767 3,040 4,818

Flares 15 20 0 1 0 4

Fracturing Liquid Tanks 0 3 0 - - -

Fugitives 0 4,260 0 0 26 0

Glycol Dehydrator 0 206 0 64 3,204 128

Heaters 51 4 0 109 6 92

Loading 0 553 0 - - -

Oil Tanks 0 291 0 - - -

Separators 0 161 0 - - -

Thermal Oxidizers 6 2 0 - - -

Turbines 0 0 0 - - -

Ventsa 0 187 0 0 1,317 0

Water Tanks 0 1,343 0 - - -

Drilling Rig Engines 10,345 606 2,155 4,362 246 2,511

Fracing - - - 516 81 322

Wellpad Construction - - - 141 20 61

Pneumatic Devices - - - 0 427 0

Total 29,443 20,427 2,155 12,962 8,488 7,942 a

Blowdown and completion venting emissions are combined for Haynesville emission inventory.

5.4 Discussion

This study provides an improved estimate of potential emissions from future development of natural gas resources in the Haynesville Shale. The emission inventory compiled in this study suggests that significant emissions of ozone precursors were emitted in 2011 and the level of emissions in future years will be strongly affected by the pace of development. As discussed in the results section, midstream compressor station and natural gas plant sources are the largest NOx, VOC and CO emissions contributing category for all years and for all scenarios. It is clear from Table 5-2 that for all pollutants, (1) low scenario NOx emissions (tons/day) decrease from 2011 to 2016 and remain approximately constant thereafter, (2) moderate scenario NOx emissions (tons/day) decrease from 2011 to 2016 and increase thereafter, and (3) aggressive scenario NOx emissions (tons/day) increase monotonically from 2011 to 2020.

It is important to note that there is uncertainty in a number of factors that affect this study’s forecasts of future year activity and emissions. In particular, the pace of development in the future has been estimated, but is highly uncertain and may deviate from any of the specific projections used in this study for a wide variety of reasons. However using a high and low scenario bounds projected emissions in order to provide a reasonable range of potential future emissions.

January 2013

86

6.0 NEXT STEPS

This report presents the methodology and results of the updated emissions inventory development for the Haynesville Shale. This inventory is currently being processed for use in photochemical ozone modeling in Northeast Texas. The results of the ozone modeling will be reported separately.

January 2013

87

7.0 REFERENCES

Armendariz, A., 2009. “Emissions from Natural Gas Production in the Barnett Shale Area and Opportunities for Cost-Effective Improvements”. Report prepared for Ramon Alvarez, Ph.D., Environmental Defense Fund, 44 East Avenue, Suite 304, Austin, Texas 78701. January. Accessed online December, 2012 at: http://www.edf.org/documents/9235_Barnett_Shale_Report.pdf .

Baker Hughes, 2012 Rotary Rig Count Data. Accessed online December, 2012 at: http://investor.shareholder.com/bhi/rig_counts/rc_index.cfm?showpage=na

Bar-Ilan, A., J. Grant, R. Friesen, A. Pollack, D. Henderer, D. Pring, K. Sgamma, 2008a. “Development of Baseline 2006 Emissions from Oil and Gas Activity in the Denver-Julesburg Basin” Prepared for Western Governor’s Association. Prepared by ENVIRON International Corporation, Novato, CA. Accessed online December, 2012 at: http://www.wrapair2.org/PhaseIII.aspx.

Bar-Ilan, A. R. Parikh, J. Grant, T. Shah, A. Pollack, 2008b. Recommendations for Improvements to the CENRAP States’ Oil and Gas Emissions Inventories”. Prepared for Central States Regional Air Partnership 10005 S. Pennsylvania, Ste. C, Oklahoma City.

Bar-Ilan, A., Grant, J., R. Parikh, Pollack, A., Henderer, D., Pring, D., Sgamma, K., 2009a. “Development of 2012 Oil and Gas Emissions Projections for the Piceance Basin”; Prepared for Western Governor’s Association by ENVIRON International Corp., Novato CA. Accessed online December, 2012 at: http://www.wrapair2.org/PhaseIII.aspx

Bar-Ilan, A., J. Grant, R. Parikh, A. Pollack, D. Henderer, D. Pring, K. Sgamma, 2009b. “Development of Baseline 2006 Emissions from Oil and Gas Activity in the North San Juan Basin” Prepared for Western Governor’s Association. Prepared by ENVIRON International Corporation, Novato, CA. http://www.wrapair2.org/PhaseIII.aspx

Bar-Ilan, A., R. Friesen, R. Parikh, J. Grant, A. Pollack, D. Henderer, D. Pring, K. Sgamma, 2009c. “Development of Baseline 2006 Emissions from Oil and Gas Activity in the Uinta Basin” Prepared for Western Governor’s Association. Prepared by ENVIRON International Corporation, Novato, CA. http://www.wrapair2.org/PhaseIII.aspx

Bar-Ilan, A., R. Friesen, R. Parikh, J. Grant, A. Pollack, D. Henderer, D. Pring, K. Sgamma, 2009d. “Development of Baseline 2006 Emissions from Oil and Gas Activity in the South San Juan Basin” Prepared for Western Governor’s Association. Prepared by ENVIRON International Corporation, Novato, CA. http://www.wrapair2.org/PhaseIII.aspx

Bar-Ilan, A., R. Friesen, R. Parikh, J. Grant, A. Pollack, D. Henderer, D. Pring, K. Sgamma, 2010. “Development of Baseline 2006 Emissions from Oil and Gas Activity in the Wind River Basin” Prepared for Western Governor’s Association. Prepared by ENVIRON International Corporation, Novato, CA. http://www.wrapair2.org/PhaseIII.aspx

http://www.edf.org/documents/9235_Barnett_Shale_Report.pdf

http://investor.shareholder.com/bhi/rig_counts/rc_index.cfm?showpage=na

http://www.wrapair2.org/PhaseIII.aspx






January 2013

88

Bar-Ilan, A., R. Friesen, R. Parikh, J. Grant, A. Pollack, D. Henderer, D. Pring, K. Sgamma, 2011. “Development of Baseline 2006 Emissions from Oil and Gas Activity in the Powder River Basin” Prepared for Western Governor’s Association. Prepared by ENVIRON International Corporation, Novato, CA. http://www.wrapair2.org/PhaseIII.aspx

Bar-Ilan, A., R. Friesen, R. Parikh, J. Grant, A. Pollack, D. Henderer, D. Pring, K. Sgamma, 2012. “Development of Baseline 2006 Emissions from Oil and Gas Activity in the Greater Green River Basin” Prepared for Western Governor’s Association. Prepared by ENVIRON International Corporation, Novato, CA. http://www.wrapair2.org/PhaseIII.aspx

Boyer, 2012. Personal communication with Doug Boyer, Texas Commission on Environmental Quality. August.Chemical and Engineering News, 2012. “Sasol moves ahead with massive Louisiana chemical and fuels project”. December 12. www.cen-online.org.

Energy Information Administration (EIA). 2011a “What is shale gas and why is it important?” Accessed online December, 2012 at: http://www.eia.gov/energy_in_brief/about_shale_gas.cfm.

Energy Information Administration (EIA). 2011b “Haynesville surpasses Barnett as the Nation's leading shale play”. Accessed online December, 2012 at: http://www.eia.gov/todayinenergy/detail.cfm?id=570

Energy Information Administration (EIA). 2011c “Review of Emerging Resources: U.S. Shale Gas and Shale Oil Plays“. Accessed online December, 2012 at: http://www.eia.gov/analysis/studies/usshalegas/

Energy Information Administration (EIA). 2012, Presentation: “Emerging Oil & Gas Supplies: Future Prospects for Oil & Gas Production”. Accessed online December, 2012 at: http://www.eia.gov/pressroom/presentations/staub_06272012.pdf

ENVIRON, 2009. “Development of Emissions Inventories for Natural Gas Exploration and Production Activity in the Haynesville Shale”; Prepared for East Texas Council of Governments by ENVIRON International Corp., Novato CA, August.

EPA, 1995. “AP 42, Fifth Edition Compilation of Air Pollutant Emission Factors, Volume 1: Stationary Point and Area Sources”; January. Accessed online December, 2012 at: http://www.epa.gov/ttn/chief/ap42/

EPA, 2004. “Convert Gas Pneumatic Controls to Instrument Air, Lessons Learned”. EPA-430-B-04-003, February.

EPA, 2008. “User’s Guide for the Final NONROAD2005 Model”. Assessment and Standards Division Office of Transportation and Air Quality, United States Environmental Protection Agency, Prepared with assistance from Cimulus, Inc. and ENVIRON International Corp. December.



http://www.eia.gov/energy_in_brief/about_shale_gas.cfm

http://www.eia.gov/todayinenergy/detail.cfm?id=570

http://www.eia.gov/analysis/studies/usshalegas/

http://www.eia.gov/pressroom/presentations/staub_06272012.pdf

http://www.epa.gov/ttn/chief/ap42/

January 2013

89

EPA, 2009. “Suggested Nationwide Average Fuel Properties”. Assessment and Standards Division Office of Transportation and Air Quality, United States Environmental Protection Agency. April.

ERG, 2009. “Drilling Rig Emission Inventory for the State of Texas.” Final Report. Prepared for

Texas Commission on Environmental Quality (TCEQ) by Eastern Research Group. Accessed

online December, 2012 at:

http://www.tceq.state.tx.us/assets/public/implementation/air/am/contracts/reports/ei/58207

83985FY0901-20090715-ergi-Drilling_Rig_EI.pdf

ERG, 2012. “Condensate Tank Oil and Gas Activity.” Final Report. Prepared for Texas Commission on Environmental Quality by Eastern Research Group, October, 2012.

Groundwater Resources Council and ALL Consulting, 2009. “Modern Shale Gas Development in the United States: A Primer”. Prepared for U.S. Department of Energy Office of Fossil Energy and National Energy Technology Laboratory. April.

Kemball-Cook, S., J. Johnson, E. Tai, M. Jimenez, G. Mansell, and G. Yarwood. 2008. “Modeling Northeast Texas Ozone for May-June 2005”. Prepared for the East Texas Council of Governments. 3800 Stone Road, Kilgore, TX, 75662. August 31.

Kemball-Cook, S., J. Jung, W. Santamaria, J. Johnson, E. Tai, M. Jimenez, and G. Yarwood. 2010. “Modeling Northeast Texas Ozone of 2005 and 2012”. Prepared for the East Texas Council of Governments. 3800 Stone Road, Kilgore, TX, 75662. March 31. http://www.netac.org/UserFiles/File/NETAC/5_25_10/TC/Enclosure_TC2.pdf.

Louisiana Department of Natural Resources (LNDR) 2012a. “Haynesville Shale Well Activity by Month, December 4th 2012” December 17,2012. Accessed online at: http://dnr.louisiana.gov/assets/OC/haynesville_shale/haynesville_monthly.pdf.

Louisiana Department of Natural Resources (LNDR) 2012b. “Haynesville Shale Scout Report”, August,2012.Nossiter, A., 2008. “Gas Rush Is On and Louisianans Cash In,” New York Times, August 2008. Accessed online December, 2012 at: http://www.nytimes.com/2008/07/29/us/29boom.html?_r=1&oref=slogin.

Louisiana Department of Natural Resources (LNDR) 2012c Haynesville Shale website. Accessed online December, 2012 at: http://dnr.louisiana.gov/haynesvilleshale/

Louisiana Department of Natural Resources (LNDR) 2012d, Strategic Online Natural Resources Information System. Accessed online December, 2012 at: http://sonris.com/

Stoeckenius, T., and G. Yarwood, 2004. “Conceptual Model of Ozone Formation in the Tyler/Longview/Marshall Near Nonattainment Area”. Prepared for East Texas Council of Governments. January, 2004.

Texas Commission on Environmental Quality (TCEQ). 2010. “Emissions Inventory Processes, Recent Research and Improvements, and the Barnett Shale Special Inventory.” Whitten, Miles T. October 16. Accessed online December, 2012 at:

http://www.tceq.state.tx.us/assets/public/implementation/air/am/contracts/reports/ei/5820783985FY0901-20090715-ergi-Drilling_Rig_EI.pdf

http://www.tceq.state.tx.us/assets/public/implementation/air/am/contracts/reports/ei/5820783985FY0901-20090715-ergi-Drilling_Rig_EI.pdf

http://www.netac.org/UserFiles/File/NETAC/5_25_10/TC/Enclosure_TC2.pdf

http://www.nytimes.com/2008/07/29/us/29boom.html?_r=1&oref=slogin

http://dnr.louisiana.gov/haynesvilleshale/

http://sonris.com/

January 2013

90

http://www.tceq.texas.gov/assets/public/implementation/air/ie/pseiforms/10162010arlington.pdf

Texas Commission on Environmental Quality (TCEQ). 2011a. “Barnett Shale Special Inventory, Phase Two. Summary Data Information.” Accessed online December, 2012 at: http://www.tceq.texas.gov/assets/public/implementation/air/ie/pseiforms/-summarydatainfo.pdf

Texas Commission on Environmental Quality (TCEQ). 2011b. “Barnett Shale Area Special Emissions Inventory: Phase I.” Accessed online at: (www.tceq.texas.gov).

Texas Railroad Commission (TRRC) 2012a Haynesville Shale website. Accessed online December, 2012 at: http://www.rrc.state.tx.us/bossierplay/index.php

Texas Railroad Commission (TRRC) 2012b. Oil and Gas Data Query: General Production. Accessed online December, 2012 at: http://webapps2.rrc.state.tx.us/EWA/productionQueryAction.do

Wall Street Journal, 2008. “Chesapeake, Plains Set to Tap Gas Field”. Accessed online December, 2012 at: http://online.wsj.com/article/SB121501262685723169.html

Wall Street Journal, 2009. “Chesapeake Continues Gas-Production Increase”. Accessed online December, 2009 at: http://online.wsj.com/article/SB124899172528895117.html/.



http://www.tceq.texas.gov/assets/public/implementation/air/ie/pseiforms/-summarydatainfo.pdf

http://www.tceq.texas.gov/assets/public/implementation/air/ie/pseiforms/-summarydatainfo.pdf

http://www.tceq.texas.gov/

http://www.rrc.state.tx.us/bossierplay/index.php

http://webapps2.rrc.state.tx.us/EWA/productionQueryAction.do

http://online.wsj.com/article/SB121501262685723169.html

http://online.wsj.com/article/SB124899172528895117.html/

January 2013

APPENDIX A Letter and Questionnaire Sent to Haynesville Shale Producers

January 2013

A-1

Appendix A. Letter and Questionnaire Sent to Haynesville Shale Producers

May 10, 2012

Producer Address, Dear , Since 1996, Northeast Texas Air Care (NETAC) has worked to study our local ozone air quality issue and to develop voluntary programs that improve air quality while protecting our regional economy. As co-chairs of NETAC, we have seen on a first-hand basis the success of these efforts which have allowed us to avoid federal intervention in the form of an ozone non-attainment designation despite more stringent ozone standards. We are contacting you seeking general information regarding equipment use at your company’s natural gas wells in the Haynesville Shale. This information will be used to update our existing inventory of emissions from exploration and production activities in the Haynesville Shale. .

The development of natural gas resources within the Haynesville Shale is an important driver of local economic growth, but also generates emissions of nitrogen oxides and volatile organic compounds that lead to ozone formation. A projected increase in emissions of ozone-forming compounds in the Haynesville Shale at a time when Northeast Texas is near to non-attainment of the National Ambient Air Quality Standard for ozone is an issue that NETAC has determined should be assessed.

NETAC has retained ENVIRON to improve the Haynesville emission inventory. ENVIRON is surveying producers active in the Haynesville in order to obtain the most accurate information available on equipment use in the field. Producers are the only source of accurate information on well site equipment use, and this data is vital to the development of an accurate inventory.

ENVIRON will use the information provided to estimate present and future emissions from exploration & production and gathering/transmission operations associated with natural gas development of the Haynesville Shale in Northwest Louisiana and Northeast Texas. The emission inventory will then be used in a computer model of the atmosphere to forecast the impacts of development in the Haynesville Shale on ozone air quality in Northeast Texas. In the absence of actual information regarding Haynesville Shale and equipment use (e.g. time required to drill a Haynesville well), ENVIRON will be obliged to make worst-case assumptions as they characterize emissions from exploration and production sources.

We would greatly appreciate it if you would complete as many questions as possible and return the questionnaire to ENVIRON by April 30, 2012. If you are able to fill out only a part of this form, please return it anyway - all information provided will be valuable in developing the inventory.

January 2013

A-2

If you have any questions about the questionnaire or how the information collected will be used, please contact Shagun Bhat at ENVIRON at (713) 470-2648. If you would like to learn more about NETAC, please see http://www.netac.org/ or contact Rick McKnight at (903-984-8641), the environmental manager for the East Texas Council of Governments.

Sincerely, Signature

http://www.netac.org/

January 2013

A-3

Questionnaire Sent to Haynesville Shale Producers

Overview Northeast Texas Air Care (NETAC) Emissions Inventory Improvement for Emissions from Natural Gas Exploration and Production Activities in the Haynesville Shale

Background: Northeast Texas Air Care (NETAC) is a voluntary, cooperative association of local governments and industries that was formed to fill the need of a more organized and comprehensive approach to improving air quality based on regional needs (http://www.netac.org). NETAC has developed an emission inventory for the Haynesville Shale which will be used for ozone air quality planning. NETAC has retained ENVIRON to improve the Haynesville emission inventory. ENVIRON is surveying producers active in the Haynesville in order to obtain the most accurate information available on equipment use in the field. Producers are the only source of accurate information on well site equipment use, and this data is vital to the development of an accurate inventory.

Filling Out the Survey Form: The survey should take no more than 30 minutes to complete. In the tabs numbered 1-3, you will find a brief questionnaire asking about equipment present at your company's typical Haynesville well. Tab 1 requests information regarding the amount of time required to drill a Haynesville well and the types of production equipment present at a typical well. Tab 2 asks several questions about the field-wide prevalence of equipment. Tab 3 requests more detailed information on typical well site equipment and activity.

The findings of this project are critical to ensuring that Haynesville Shale emissions are correctly represented in the Texas SIP modeling. If you are only able to fill out a part of this form, please return it anyway - all information provided will be valuable in improving the inventory. Thank you for your time and cooperation in filling out the survey form.

ENVIRON Contact Information:

Mr. Shagun Bhat

ENVIRON International Corporation

Address: 10333 Richmond Avenue, Suite 910, Houston, TX 77042

Phone: 713.470.2648

Email: [email protected]

http://www.netac.org/

January 2013

A-4

Tab 1: TYPICAL SITE INFORMATION

TYPICAL SITE INFORMATION

Instructions

Please provide complete information in 'green' cells.

Drilling Information - Please answer the following for drilling at your

company's typical Haynesville well

Production equipment for your company's typical Haynesville well -

Please indicate whether each type of equipment is present and if present,

number of units :

Yes No

If Yes, please

provide number of

units

Compressor EnginesHeater/Treater

Condensate Tanks

Produced Water Tanks

Other (please specify)

Construction Equipment Information - Please answer the following for

your company's typical Haynesville well

What is the total engine time to drill a

typical well in the Haynesville Shale area?

How much area is cleared for a typical

well pad? (acres)

What is the length of the access road to

the well pad? (miles)

January 2013

A-5

Tab 2: FIELD-WIDE INFORMATION

FIELD-WIDE INFORMATION

Instructions

Please provide complete information in 'green' cells

Do you have well pads with both Haynesville and non-Haynesville If yes, please provide the number of such well pads?

Compressor Information - Please answer the following for your company's

Haynesville wells

What is the approximate percentage of your company's Haynesville

Shale wells with a tank battery?

What approximate percentage of your

company's compressor engines in the

Haynesville Shale area are Rich-burn?

What approximate percentage of your

company's Haynesville wellheads have a

compressor?

January 2013

A-6

Tab 3: DETAILED SITE INFORMATION

DETAILED SITE INFORMATIONInstructionsPlease provide complete information in 'green' cells

Drilling Information - Please answer the following for drilling at your


List Equipment Used to

Clear PadConstruction Equipment Information - Please answer the following for your


What kind of equipment is used to clear your company's

typical Haynesville well pad area - Please specify

horsepower and type of equipment and time it takes to

clear a typical well pad area.

How long does it take to clear your company's typical

Haynesville well pad area?

What are the total engine horsepower-hours to drill your

company's typical well in the Haynesville Shale area?

(horsepower-hours per well)

What is the amount of diesel fuel consumed to drill your

company's typical well in the Haynesville Shale area?

(gallons per well)

What is the typical Tier rating of the diesel engines used

during drilling in the Haynesville Shale area?

January 2013

A-7

Tab 3: DETAILED SITE INFORMATION (CONTINUED)

DETAILED SITE INFORMATION

Heater/Treater Information - Please answer the following for a heater at your

company's typical Haynesville wellFiring Rate

(MMBTU/hr)

Operating

During

Summer?

Typical Yearly Working Hours

(hours per year)

Compressor Information - Please answer the following for a compressor at

your company's typical Haynesville wellMake Model No. Typical Load Factor (%)

Please add

information on

controls, if any

Condensate Tanks - Please answer the following for tanks at your company's

typical Haynesville well

What is a typical configuration for condensate and produced storage

tanks at a well site - number, type (fixed/floating), size (height and diameter),

service (produced water or condensate) ?

Has E&P Tank, HYSYS, Promax or any other simulation software been

used to estimate hydrocarbon emissions from condensate storage tanks?

If yes, please provide a sample run summary.If no, please provide condensate compositional analysis

Has flash gas analysis been performed for the condensate produced?If yes, please provide the flash gas analyses.

Has compositional analysis been performed for produced gas?

If yes, please provide the compositional analyses.

Production Equipment Information - If the equipment listed below is present

at your company's typical Haynesville well, please provide additional detail

Date post:	21-Jan-2021
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

East Texas Council of Governments · Shagun Bhat Amnon Bar-Ilan Sue Kemball-Cook Greg Yarwood...

Documents