Robert Margolis and Paul Denholm NREL Presented at the DOE Workshop on Estimating the Benefits and Costs of Distributed Energy Technologies Sept. 30, 2014
Estimating the Costs and Benefits of Avoided Generation Capacity
Estimating Capacity Value Involves 2 Steps
1) Determine capacity credit, i.e., the fraction of a DET’s capacity that adds to system reliability.
2) Translate capacity credit into a monetary value, i.e., capacity value.
Note: In the literature capacity credit and capacity value are often used interchangeably. Here following Mills and Wiser (2012) we use capacity credit to refer to physical capacity and capacity value to refer to economic value (measured in $ or $/MW).
2
1) Methods to Calculate Capacity Credit
Name Description Tools Required
Capacity factor approximation using net load
Examines output during periods of highest net demand Spreadsheet
Capacity factor approximation using loss of load probability (LOLP)
Examines output during periods of highest LOLP Spreadsheet
Effective load-carrying capacity (ELCC) approximation (Garver’s Method)
Calculates an approximate ELCC using LOLPs in each period
Spreadsheet
Full Effective load-carrying capacity (ELCC) calculation
Performs full ELCC calculation using iterative LOLPs in each period
Dedicated tool
3
Capacity Factor Approximation Methods Basic Method: • Examine generator output/
capacity factor during periods of high net load or periods of highest risk
• Choice of peak period (top 100 hours, top 1% etc) can significantly influence results
Pros/Cons • Very easy, useful for rough
estimates • Requirements are only load,
DET profiles and a spreadsheet
• Still somewhat common although decreasingly so…
0
2,000
4,000
6,000
8,000
10,000
12,000
14,000
0
100
200
300
400
500
600
700
0 6 12 18 24
Base
Cas
e N
et L
oad
Gen
erat
or O
utpu
t (M
W)
Hour
Limited EnergyResource(Dispatched)
PV
LOAD
Peak Hour on this day
Simple Example
But period examined is important
4
Examples of Peak Periods used for Capacity Value Approximation Methods in the United States
Rogers and Porter (2012), “Summary of Time Period-Based and Other Approximation Methods for Determining the Capacity Value of Wind and Solar in the United States.” NREL Subcontract report. Available at http://www.nrel.gov/docs/fy12osti/54338.pdf
5
1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 10
20
30
40
50
60
70
80
90
100
SM
Nor
mal
ized
Cap
acity
Val
ue (%
)
ELCC Top Weighted Top LOLP Top Load
CF Approximation Methods may not Fully Measure Contribution to Resource Adequacy
• DET generation-load relationship is only part of the equation
• Capacity factor, even during peak periods, won’t capture annual risk profile
• Improvement is to use CF during period of high risk (high LOLP periods)
• All CF based approaches are inherently limited
6
Comparison of Capacity Factor Methods for CSP with Increasing
Levels of Storage
Reliability (ELCC) Based Approach
How does ELCC work? • Holds the system at constant annual risk level
with/without the generator of interest (wind, solar, etc.) • Utilizes reliability/production simulation model
– Hourly loads – Generator characteristics – DET generation pattern (hourly for >= 1 year) time-
synchronized with load – Calculates hourly LOLP (loss of load probability)
• The hourly LOLP calculation finds high-risk hours: risk can be caused by – Peak loads – Unit unavailability (planned maintenance) – Interchange and hydro schedules/availability
• Most hours/days have LOLP=0 so are discarded: only high-risk/peak hours remain in the calculation of ELCC
• Conventional units ELCC is function of FOR (forced outage rate)
7
Steps to Calculating ELCC
A. Develop benchmark system excluding DET that meets the target reliability level (1 day in 10 years is a common target)
B. Add DET and rerun model, noting annual reliability (this is the DET case)
C. Incrementally increase load until annual reliability matches the benchmark system case
The capacity of the added load is the ELCC of the DET
8
A) Benchmark System B) with DET
0.12
0.11
0.10
0.09
0.08
0.07
0.06
Loss
of L
oad
Expe
ctat
ion
(day
s/ye
ar)
11.0x10 3 10.5 10.0 9.5 9.0 8.5 8.0 Load (GW)
Original Reliability Curve Reliability Curve after Adding New Generation Target Reliability Level
<-400 MW->
Each generator added to the system helps increase the load that can be supplied at all reliability levels
With new DET reliability has increased
A
B
9
C) Return to Target Reliability Level
0.12
0.11
0.10
0.09
0.08
0.07
0.06
Loss
of L
oad
Expe
ctat
ion
(day
s/ye
ar)
11.0x10 3 10.5 10.0 9.5 9.0 8.5 8.0 Load (GW)
Original Reliability Curve Reliability Curve after Adding New Generation Target Reliability Level
<-400 MW->
Each generator added to the system helps increase the load that can be supplied at all reliability levels A
B
C
ELCC
10
Increase load to original risk target
ELCC Data Requirements • Hourly DET
generation and load data must be from the same year
• State of the art is to use meso-scale weather models or actual DET production data (same as integration studies)
• Want to preserve underlying correlations between DETs and load with weather, etc.
Should use multiple years of data
Minnesota 20% Wind Integration StudyWind Capacity Value (ELCC) by Penetration
05
10152025
2003 2004 2005
Year
Perc
ent o
f Win
d R
ated
Cap
acity
15%20%30%
11
• Generator capacity, forced outage rates • Maintenance schedules • Hydro schedules • Transaction schedules • Transmission limits • Calculation over large part of the interconnection
with transmission • All of these will have some influence on
LOLP and therefore VG ELCC
2) Translating Capacity Credit to $ Value
12
Name Description Tools Required
Simple avoided generator (CT)
Assumes DG avoids construction of a new CT None
Weighted avoided generator
Assumes DG avoids a mix of generators based on avoided fuel None
Capacity market value
Uses cost of capacity in restructured markets None
Screening curve Uses system load and generation data to estimate avoided generation mix based on capacity factor
Spreadsheet
Complete valuation of DET versus alternative technologies
Estimates the type or mix of generators avoided in subsequent years using a capacity-expansion model
Detailed capacity-expansion model
DETs May Avoid a Mix of Generation Types
Bottom line…. • Lower Bound Value – DET avoids only CTs • Upper Bound Value– DET avoids CCs
13
4000
5000
6000
7000
8000
9000
10000
11000
12000
13000
14000
0 2000 4000 6000 8000 10000
Load
(MW
)
No DG
With DGPV
Capacity Value Depends on Level of Penetration
Most studies indicate that by 10% energy penetration capacity credit and capacity value of DG PV is very low
14
Could see Interactions between DETs
0
10,000
20,000
30,000
40,000
50,000
60,000
0% 2% 6% 10%
PV Penetration and Hour
Gen
erat
ion
(MW
)
PVCTPSHydroCCImportsCoalNuclearWindGeo
Base 2% 6% 10% (no PV)
Here PV narrows the peak, which could ease the ability of demand response, hydro, storage, etc. to provide capacity
15
Additional Issues
• Fundamental relationship between weather and solar/wind profiles means need for multiple years of data/analysis
• Transmission – capacity sharing would impact DET capacity value
• As penetration of DETs increases, could see interactions between technologies.
• ELCC addresses the issue of sufficient planning reserves—is there a need for a flexibility assessment? – Consider ramping needs over relevant time scales – Contribution of various sources of flexibility (including DET
control) – Probabilistic assessment of system ramping capability (effective
ramping capability)
16