Technical guides for owner/manager of an air conditioning system: volume 7

A benchmarking guide adapted to air conditioning based on electricity bills

1

AustriaAustrian Energy Agency

BelgiumUniversité de Liège

ItalyPolitecnico di Torino

PortugalUniversity of Porto

AustriaAustrian Energy Agency

AustriaAustrian Energy Agency

BelgiumUniversité de Liège

BelgiumUniversité de Liège

ItalyPolitecnico di Torino

ItalyPolitecnico di Torino

PortugalUniversity of Porto

PortugalUniversity of Porto

SloveniaUniversity of Ljubljana

UKAssociation of Building

Engineers

BRE (Building Research Establishment Ltd)

Welsh School of Architecture

SloveniaUniversity of Ljubljana

SloveniaUniversity of Ljubljana

UKAssociation of Building

Engineers

UKAssociation of Building

Engineers

BRE (Building Research Establishment Ltd)

BRE (Building Research Establishment Ltd)

Welsh School of Architecture

Welsh School of Architecture

Eurovent-CertificationEurovent-Certification

Team

France (Project coordinator)Armines - Mines de Paris

France (Project coordinator)Armines - Mines de Paris

Authors of this volume Jérôme ADNOT (Armines, France) Daniela BORY (Armines, France) Maxime DUPONT (Armines, France) Roger HITCHIN (BRE, UK)

The sole responsibility for the content of this publication lies with the authors. It does not represent the opinion of the European Communities. The European Commission is not

responsible for any use that may be made of the information contained therein.

2

Objectives of the guide The objective of this guide is mainly to help building owners to detect changes in energy consumption of their air conditioning plant as a first step in identifying and correcting faults. Even in the absence of obvious faults , building owners should monitor energy consumption and be alert for early signs of problems. This will enable them to react promptly through behavioral changes, operational changes (system adjustments), investment in component replacement or in extreme cases with a complete retrofit. When changes have been introduced, monitoring allows the effectiveness and actual payback times to be assessed for future use. In order to help the managers an action plan has been defined: a complete scheme of the plan is reported in the next page and it is detailed all along the document. Moreover, some example of ratios of energy consumption are presented in the annex as indicative values that should be considered for their order of magnitude more than the value itself.

3

4

1st Part: Basic benchmarking of energy consumption

Collect activity indicators that might

justify variations in electricity consumption

Note any changes other that climate (process, adjustments) from one time

period to another

Collect electricity bills (prefer monthly to annual basis to get more accuracy)

Draw the diagram: time period on the X-axis and kWh/time period on the Y-

axis

Try to explain variations in consumption thanks to previous

changes or activity indicators

If variations are unexplainable, call a professional (energy supplier, ESCO

etc.) for an energy audit

Try to continue that follow-up of energy consumption even if no

problem is detected

2nd Part: Benchmark typical energy ratios

Choose main activity indicators that

justify variations in electricity consumption

Calculate electricity consumption on the time period (the same as the one used in

statistics)

For each activity indicator and the same time period, divide energy consumption

by the indicator

Compare each ratio from one time period to another

Try to explain possible variations

(climate, adjustments etc…)

Compare each ratio to statistics

Try to explain possible differences (size, climate, adjustments etc…)

If variations/differences are

unexplainable, call a professional (energy supplier, ESCO etc.) for an

energy audit

Try to continue that benchmarking of energy ratios even if no problem is

detected

3rd Part: climate based energy signature

Determine the best time period

(monthly/weekly for DD, daily for average temperature)

Determine the best temperature reference (usually 18°C in winter, from 14°C to 24°C

in summer)

Measure/get average daily temperature or calculate/get degree-days for each time

period

Separate cooling degree-days from heating degree-days

Measure /get from the bills electricity

consumption on each time period

Draw the diagram for each time period: average temperature or DD on the X-axis

and kWh on the Y-axis As soon as a season/year is complete,

determine the (linear) regression between consumption and climate parameters

Try to find the reasons why a new dot is

outside the scatter – If differences are too important, too frequent and

unexplainable, call a professional (energy supplier, ESCO etc.) for an energy auditCalculate climate independent energy

ratios (slope of the regression curve) in kWh/m2.DD

Compare to statistics – Try to explain

possible differences (size, activity, adjustments etc…)

4th Part: multi-parameter energy signature

Choose main activity and climate indicators that justify variations in

electricity consumption

Determine the best time period (monthly, weekly, daily)

Measure/get every chosen indicators for

each time period

Measure/get from bills the electricity consumption of each time period

If enough dots are available, establish the linear regression (E=Σai.Pi + b)

between electricity consumption E and every parameters Pi

Try to find the reasons why a new dot is outside the scatter – If differences are

too important, too frequent and unexplainable, call a professional

(energy supplier, ESCO etc.) for an energy audit

Calculate energy ratios by extracting

multiplier coefficients ai

Compare to statistics – Try to explain possible differences (size, activity,

adjustments etc…)

Try to continue the energy signature even if no problem is detected

PRELIMINARY ACTIONSDetermine the electricity uses included in the bill in general

Collect electricity bills (prefer monthly to annual basis to get more accuracy) Determine parameters having effect on electricity consumption in general

ACTION PLAN

1. Pre-requisite Energy uses having effect on the electricity bill of a building Any building owner must be aware that, for more accuracy, direct measurements on individual items of equipment (chiller for example) or energy use (lighting for example) - which may need to be organized by himself – are preferable overall utility energy bills. The latter combine energy consumption by numerous end-uses and may relate to time periods that are not ideal for monitoring purposes. The total electricity bill mainly includes the consumption of lighting, ventilating and small power appliances (kitchen or office equipments, HI-FI etc…) and sometimes process loads such as air compressors, pumps, fans etc…Air conditioning, is included because non-electric cooling is rare. Building heating and domestic water heating can also be provided by the Joule effect (direct electric) or with a heat pump. However these two activities often use fossil fuels and are then accounted in a separate bill. Interpretation of electricity bills requires the listing of each type of electrical appliances whose consumption is included in the energy bill. When possible, try to evaluate the electrical power they absorb in operation by looking at their electrical plate or documentation (see Table 1).

Appliances type Number of equipments

Electric unit power

Operation hours Energy (kWh)

Computers 40 300 4380 13140 Printers type A 10 400 500 2000

(20 stand by) 1000 200

Bulbs type X 50 60 5000 3000 Bulbs type Y 10 100 5000 5000 Neon type Z 70 40 1540 616

Electric heaters 10 1200 2500 30000 Etc. …. … … …

Total energy (kWh) xxx

Table 1 An example of listing the electric appliances and their nominal power.

It is important also to estimate the number of hours they operate during the period covered by the meter readings.. Absorbed power and time of operation can be used to estimate the energy consumption. The total amount of energy calculated should be compared to the actual measured consumption in order to verify the consistency of the list and its exhaustiveness. Keep in mind that the analysis of air conditioning performances using global energy bills will be accurate only if the share of air conditioning energy consumption in the bill is significant. If energy consumption for cooling is submerged by that from other uses, accurate estimation will be impossible without sub-metering. Parameters having effects on air conditioning energy consumption Several parameters have effects on air conditioning energy consumption and more generally on all thermal energy consumption. It is possible to distinguish four types:

5

- Building parameters are intrinsic to the construction of the building. The building shell thermal characteristics, the glazing surface, heated/cooled areas and their location are part of them. They were fixed by the first owner of the building and are difficult for following building owners to change because of the costs of retrofits and relatively long payback times.

- Policy parameters depend only on the current building owner’s decisions and his will to save energy. Equipment choices and investments, operational parameters such as temperature and humidity set points (if building centralized), the time of operation or maintenance and follow-up policies are among them. The sensitivity of energy consumption to these parameters is really important.

- Behavioral parameters depend on the occupants’ choices. Operational parameters such as temperature and humidity set points (if room localized) or the natural uneconomic behavior of occupants are part of them. Their influence on energy consumption can be large although the duration of the “good practice” behavior can be short.

- Activity parameters depend mostly on the use of each space . These are largely determined by the business needs of the organization. They have an important influence on energy consumption but a building owner or energy manager cannot usually change them.

Interpretation requires the identification of the most important parameters affecting each type of energy consumption. Special attention must be paid to causes that might be improved by the building owner. These are then the preliminary actions to develop in order to begin the analysis of the energy consumption of the building and are necessary to all the processes developed in the following paragraphs.

PRELIMINARY ACTIONS Determine the electricity uses included in the bill in general

Collect electricity bills (prefer monthly to annual basis to get more accuracy)Determine parameters having effect on electricity consumption in general

Figure 1: Preliminary part of ACTION PLAN

6

2. Basic benchmarking of energy consumption With such numerous parameters, it is clear that the energy consumption of buildings can differ strongly. The first part of this guide is dedicated to the simple follow-up of energy consumption of one building from one time period to another. Most of owners only focus on energy bills at the building level looking almost only at the amount of money they pay, but several factors should be considered when comparing bills such as energy prices (that vary in general form one year to another), climate effects, activity and changes in the building structure, operation etc. Thinking about energy uses can bring however a lot of additional information. After listing energy uses and influent parameters, it is important for a building owner to determine the optimal measurement frequency (less than an hourly, hourly, daily, weekly, monthly, yearly etc…). The objective is to maximize the information available while minimizing the quantity of measurements and associated costs. Again, when certain parameters (climate for example) vary a lot, it is relevant to reduce the time period to get more accuracy. Prefer therefore monthly billing to yearly billing for thermal uses. If it is not possible, try to make your own measurement using additional metrology (general meter, sub-meter, portable metrology). As soon as consumption of at least two time periods for the same building are available, you can begin the comparison. The objective of such a follow-up is not only to be aware of variations in energy consumption but also to look at parameters responsible of such variations. Logically if thermal uses (pure cooling system, joule effect heating or reversible heat-pump) are not negligible in the bill in comparison to process uses, electricity consumption must be maximal in summer (July or August) and in winter (December, January or February). Therefore, detect possible problems on energy consumption magnitudes and variations with time by trying to explain the bill through the main influent parameters you found previously. Once the responsible parameters are underlined, think to possible actions that would allow optimizing energy consumption1 or call a professional to assist you and perform a detailed energy audit. Observe for example the variations of annual electricity consumption in some banking agencies (Figure 2). Electricity uses (office equipments, cash dispensers, lighting, HVAC) are known and unlikely to change unless in terms of quantity especially for cash dispensers. Opening hours are constant from one year to another in a bank agency so that electrical appliances operate for the same duration each year. On these examples, the area of the bank agencies and the staff are unchanged from one year to another. The only parameters that can explain such variations could be: activity variations of the agency (customers?), process changes (quantity, type?), operational changes (temperature set-points, operating hours?) or climate variations (winter, summer or both?). In that case, the agency manager is finally the best person able to determine the origins of these variations, to know if they are normal or not and to decide to carry out actions.

1 You can consult for a full list of Energy conservation opportunities the ECOs list on the AuditAC web site and the AuditAC technical guide 5 - ECOs for AC auditors.

7

Figure 2: Examples of variations in annual electricity consumption in some banking agencies.

Then observe for example the variations of monthly gas and electricity consumption in a service building in Portugal. Of course the use of gas is related only with the winter season and because the building is air conditioned all along the summer season, monthly consumptions are related to the weather.

Figure 3: Examples of variations in annual gas and electricity consumption in a library in

Portugal2.

2 AuditAC Case studies brochure: Case studies: Portugal, n°3.

8

Collect ac

Note any change

Collect e

Draw the d

Try to explain v

If variations are une

Try to contin

3. Benchmark typic The lack of referencepart of that guide are is difficult, the idea ocompared in the samthe activity level, the by dividing energy cotime period by one or The first task is to choyou want to regularly strongly and frequentIndeed, smaller is thereference allows to mperiods such as a yeconsumptions. There exist an obviouand the area of the bmeters). It can be usprocesses (office or re Another obvious relaespecially for constanthe use of that indicatnot as clear (Figure maximize the informa

1st Part: Basic benchmarking of energy consumption

tivity indicators that might justify variations in electricity consumption

s other that climate (process, adjustments) from one time period to another

lectricity bills (prefer monthly to annual basis to get more accuracy)

iagram: time period on the X-axis and kWh/time period on the Y-axis

ariations in consumption thanks to previous changes or activity indicators

xplainable, call a professional (energy supplier, ESCO etc.) for an energy audit

ue that follow-up of energy consumption even if no problem is detected

Figure 4: First part of ACTION PLAN

al energy ratios

often leads us to think that everything is OK. The objectives of the second to allow building owners to compare their energy consumption. Although it f benchmarking is then to try to build “universal” indicators that can be e activity sector whatever the building. To reach that independency from best way is to calculate energy ratios (kWh/activity parameter/time period) nsumption of a system (building, energy use, equipment) during a certain more activity indicators.

ose the best time period on which you will calculate energy ratios. Unless follow-up your energy consumption because your activity parameters vary ly, it is useless to consider short time periods such as a week or a month. reference more the climate dependence is relevant, so to choose monthly

itigate punctual peak or minimum of temperatures. Prefer typical time ar or a season also in order to distinguish between heating and cooling

s relationship between the number of energy uses, their size and power uilding so that the most common indicator is the cooled area (in square-ed in every air-conditioned spaces especially buildings without specific sidential buildings) where no other indicators are present.

tionship exists between the area of the building and its occupancy t occupancy buildings (no or limited numbers of external visitors) so that or should then be equivalent to the use of square-meters. In practice, it is 5) and the distinction between the two indicators allows sometimes to

tion.

9

Permanent occupancy = 0,0211 x Square-meters + 2,1101R2 = 0,7038

0

5

10

15

20

25

30

0 100 200 300 400 500 600 700 800 900 1000 1100 1200

Bank agency area (m2)

Perm

anen

t occ

upan

cy

Figure 5: relationship between area and permanent occupancy in the banking sector.

Everybody knows that the more the activity level of a building, the higher its energy consumption. Thermal uses are not an exception and energy consumption is correlated to occupancy. For example, heat and humidity productions and inlets due to temporary occupancy have to be fought against by air conditioning systems. As it is often difficult to count visitors of a building during a time period, it is better use other indicators that are necessarily such as: occupied beds in a hospitals, occupied bedrooms in hotels, students in schools, meals in restaurants, tickets sold in museums or concert-halls etc… In case of variable occupancy, these indicators should be used in addition to kWh/m2 ratios. Keep in mind that benchmarking only provides valid results (Figure 6) for the same building on several time periods or at least for the same category of buildings. Two buildings from the same activity sector may have totally different energy ratios because of their different sizes. Indeed, most processes, their intrinsic efficiencies and their control systems are often better in large-scale buildings. Moreover, these ratios do not take into account the climate influence (outdoor temperature, sun, wind etc…) on energy consumption so that comparisons of buildings in too different locations may be analyzed carefully.

0

50

100

150

200

250

300

350

400

450

500

550

600

0 100 200 300 400 500 600 700 800 900 1000 1100 1200

Bank angency area (m2)

Ener

gy ra

tio (k

Wh/

m2 .y

ear)

2003 2004 2005

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

0 5 10 15 20 25 30

Agency staff

Ener

gy ra

tio (k

Wh/

pers

on.y

r)

2003 2004 2005

Figure 6: the kWh/m².year ratio decreases with the surface (left) and the staff (right) in the banking sector.

10

If your ratios (compare several ones) are lower than other building ratios or statistics, your building seems energy-efficient and you have probably no action to carry out or investment to make. Continue however to regularly follow-up your indicators to anticipate problems. In the other case, you should probably contact a professional to consider some ways of improvements.

2nd Part: Benchmark typical energy ratios

Choose main activity indicators that justify variations in electricity consumption

Calculate electricity consumption on the time period (the same as the one used in statistics)

For each activity indicator and the same time period, divide energy consumption by the indicator

Compare each ratio from one time period to another

Try to explain possible variations (climate, adjustments etc…)

Compare each ratio to statistics

Try to explain possible differences (size, climate, adjustments etc…)

If variations/differences are unexplainable, call a professional (energy supplier, ESCO etc.)

for an energy audit

Try to continue that benchmarking of energy ratios even if no problem is detected

Figure 7 Second part of ACTION PLAN for benchmarking of energy ratios.

The EPlabel3 project has developed a simple software that records details of a building, its key descriptors and its energy use and produces a summary of the measured annual energy consumption and calculates Energy Performance indicators in units of total weighted energy per m² of internal area. The software main idea is to label the building energy performance as demanded in the European Energy Performance Building Directive, but it can be used to define reference ratios (as built or operational rating) and visually describe the energy content of the building. 4. Establish the climate based energy “signature” of the building Previous energy ratios are climate dependent so that only statistical ranges are published. Several climatic parameters (sun, wind, outdoor temperature, outdoor humidity) play an important role in energy consumption of thermal uses. The objective of that paragraph is then to determine the sensitivity of energy consumption to climate parameters. It is focused especially on the outdoor temperature because that parameter is easier to feel and measure for a building owner.

3 EPLabel: A programme to deliver energy certificates based on measured energy consumption for display in Public buildings across Europe within a harmonizing framework. http://www.eplabel.org/

11

The particular interest of such a method is to distinguish between climate dependent energy consumption (variable part) and them that are not (constant part). The building “signature” is then a curve that describes in a unique way the behavior if the building as a function of a climatic parameter. However, find a correlation between energy consumption and outdoor temperature necessarily requires a representative dot scattering, in other words much more dots. Therefore, monthly energy bills or weekly measurements should be preferred to annual energy bills. There are three possibilities concerning the temperature: either measure it, or get free daily-averaged measurements (can be found on the internet4) or pay for local hourly measurements (to national weather institute). Two climate indicators can be used in the energy signature. The first one is directly the averaged outdoor temperature of the time period. As a consequence, the correlation function is growing for cooling in summer and decreasing for heating in winter. The second indicator is degree-days that can be calculated for heating (HDD) in winter and cooling (CDD) in summer. HDD (respectively CDD) is the sum on the chosen time period (higher than a day) of the product of “the number of days during the time period for which the outdoor temperature is lower (respectively higher) than a reference value” by “the difference between the daily averaged outdoor temperature and that reference value”. HDD (respectively CDD) represent how cold (respectively hot) is a winter (respectively summer) time period. The main difficulty of using degree-days is to choose the good reference temperature. In theory, that value is the averaged temperature ensured only by the building shell without any additional heating (or cooling) system. Logically, the balance temperature in winter is different from the one in summer. For heating, the reference temperature is usually taken equal to 18°C in Europe. As heat gains can contribute to about 3°C in buildings and due to the averaged decrease of building shell thermal conductivity, the reference temperature for heating tends to rise. For cooling, the same problem exist so that in large office buildings, air conditioning system may start for an outdoor temperature of 14°C whereas in residential buildings, accounting less heat gains, the reference temperature may reach 24°C. Observe that not knowing the "proper" base temperature is, in practice, rarely a problem for detecting changes in consumption: energy consumption correlates well with "wrong" degree-days - since degree-day values to different base temperatures are strongly correlated with each other. Once you got enough dots for your energy signature, you are able to detect rapidly drifts in energy consumption due to some defaults compared to the past normal operation. It is thus better to apply this method as soon as the start-up of a new thermal process in order to have reference (without default) dots. The Figure 8 sums-up some of the typical defaults that can be detected from an energy signature analysis.

4 For example, a French website centralizes several climate data in a lot of locations in the country ahttp://www.meteociel.fr/climatologie/climato.php

12

kWh/

time

unit

CDD/time unit

kWh/

time

unit

CDD/time unit

kWh/

time

unit

CDD/time unit

kWh/

time

unit

CDD/time unit

kWh/

time

unit

CDD/time unit

kWh/

time

unit

CDD/time unit

kWh/

time

unit

CDD/time unit

kWh/

time

unit

CDD/time unit

Sun

Wind

Sun

Wind

kWh/

time

unit

CDD/time unit

kWh/

time

unit

CDD/time unit

kWh/

time

unit

CDD/time unit

kWh/

time

unit

CDD/time unit

Reference Tightness

Problems on controls or

thermal probes

Default on the supply reduction

system

Effects of other

climate

Oversizing or overcooling

Figure 8: Few problems that can be detected by the climate-based energy signature of a building.

That method works almost perfectly for heating and is often used in operation and maintenance contracts with guarantee of results. However, it is not as accurate for cooling because of the higher number and power of electrical auxiliaries, the stronger influence of humidity and solar radiation on air conditioning and the non-uniformity of temperature and humidity set points in buildings due to a lack of regulation. Once the correlation between energy consumption and degree-days or average temperature is established, climate independent energy ratios are directly available calculating the slope of the curve. It is then possible to compare your building ratios with those from any other building whatever their locations. Remember that measuring only the power absorbed by the considered thermal use (an air conditioning system for example) allows to strongly limit the offset magnitude of the curve and to better underline the effect of climate (slope) on energy consumption. Energy bills are a mix of consumption from different uses attenuating the visibility of pure thermal uses. Thus, always take care to the framework to compare ratios with statistics or between buildings.

13

3rd Part: climate based energy signature

Determine the best time period (monthly/weekly for DD, daily for average temperature)

Determine the best temperature reference (usually 18°C in winter, from 14°C to 24°C in summer)

Measure/get average daily temperature or calculate/get degree-days for each time period

Separate cooling degree-days from heating degree-days

Measure /get from the bills electricity consumption on each time period

Draw the diagram for each time period: average temperature or DD on the X-axis and kWh

on the Y-axis As soon as a season/year is complete, determine the (linear) regression between

consumption and climate parameters

Try to find the reasons why a new dot is outside the scatter – If differences are too important, too frequent and unexplainable, call a professional (energy supplier, ESCO etc.)

for an energy audit Calculate climate independent energy ratios (slope of the regression curve) in kWh/m2.DD

Compare to statistics – Try to explain possible differences (size, activity, adjustments etc…)

Try to continue the energy signature even if no problem is detected

Figure 9: Third part of ACTION PLAN for establishment and the use of the climate based energy signature of the building.

5. Establish the multi-parameter energy signature of the building We showed that energy consumption were due to a lot of parameters and underlined especially the influences of building parameters (area, staff), activity parameters (occupancy, production) and climate. However, it is impossible to analyze one parameter independently of the others. Concerning air conditioning for example, a more temperate climate than forecasts can be compensated by a much higher occupancy than in normal conditions so that the only climate or activity analysis will provide biased results. As all of these parameters have a common influence on energy consumption, it is important to analyze their effects simultaneously. The idea of that paragraph is then to find the energy signature of the building by considering not only the climate but also main activity parameters (pure occupancy, indirect occupancy, production etc…). Building parameters (surface, staff etc…) can be added in order to determine the energy signature of several similar buildings of the same activity sector (see next example). This method is exactly the same as the previous one but additional parameters are this time included into the linear regression. As for climate parameters, building and activity parameters must be measured precisely for each time period so that specific metrologies or procedures should be set-up. Again, time periods must be short in order to get more dots and then accuracy. Finally, to be able to find a correlation between consumption and all the parameters, you will

14

need at least as many time periods as parameters. For example, if the main parameters having effects on your building energy consumption E are the degree-days DD and the occupancy O in a supermarket, the regression method requires at least two time periods on which the coordinates (E, DD, O) are distinct. However, to reach higher levels of representativeness, it is better to scan a large range for every parameter. For example, in the banking sector the information available for each agency is: area S, staff Nwkr, the annual electricity consumption for three years and finally the location allowing to calculate both cooling (CDD) and heating (HDD) annual degree-days. As building parameters (area and staff) are constant for one agency in the time, they only intervene into the magnitude of consumption but not in their variations with time. On the opposite, variations of cooling and heating degree-days lead to variations in annual electricity consumption for one agency. As only three dots were available for each agency, the results of each energy signature were neither accurate nor representative. Although the approach is strictly the same, the energy signature was applied not to each agency but to the whole banking sector including every agencies in order to determine a general profile of electricity consumption. Several models were tested: influence of the area (S), of the winter climate (HDD), of the summer climate (CDD), of both winter and summer climates (H&CDD) depending on the type of heating system (joule or reversible heat-pump), of the staff (Nwkr). The Table 2 presents both results and accuracy (R2) obtained by such models. Obviously, the more the parameters, the more accurate but for comparison purposes they cannot be all kept.

Model (kWh/year) A B C R² A.HDD.S+B 0.101 11680 0.67

A.HDD.S+B.Nwkr+C 0.047 3822 4652 0.77 A.CDD.S+B 0.411 27170 0.27

A.CDD.S+B.Nwkr+C 0.139 4000 1766 0.64 Heating by Joule effect (COP=1)

A.H&CDD.S+B 0.099 11481 0.67 A.H&CDD.S+B.Nwkr+C 0.0461 3813 4700 0.77

Heating by Heat Pump (COP=2.5) A.H&CDD.S+B 0.099 11174 0.68

A.H&CDD.S+B.Nwkr+C 0.0479 3700 4740 0.77

Table 2: modeling of electricity consumption in the banking sector in France using a linear regression method

Once the multi-parameter energy signature of the building is established, drifts in energy consumption generated by defaults may be detected as soon as new dots deviate from the scatter. As for the simple climate base energy signature, energy ratios relative to each parameter may be deduced from the correlation function calculated by extracting multiplier coefficient before each variable. Eventually, the DD regression not being be enough precise or of hard interpretation, it exists other methods such as the CUSUM method5 that allow to reveal trends that are occurring in the building energy performance that are not visible through the only regression method.

5 For more detail about the CUSUM method you can look at the: “CTG004- Degree days for energy management — a practical introduction”, Carbon trust at http://www.carbontrust.co.uk/publications

15

4th Part: multi-parameter energy signature

Choose main activity and climate indicators that justify variations in electricity consumption

Determine the best time period (monthly, weekly, daily)

Measure/get every chosen indicators for each time period

Measure/get from bills the electricity consumption of each time period If enough dots are available, establish the linear regression (E=Σai.Pi + b) between electricity

consumption E and every parameters Pi

Try to find the reasons why a new dot is outside the scatter – If differences are too important, too frequent and unexplainable, call a professional (energy supplier, ESCO etc.) for an energy audit

Calculate energy ratios by extracting multiplier coefficients ai

Compare to statistics – Try to explain possible differences (size, activity, adjustments etc…)

Try to continue the energy signature even if no problem is detected

Figure 10: fourth part of ACTION PLAN for establishment and the use of the multi-parameter energy signature of the building.

16

Annex

1) Some EECCAC (Energy Efficiency and Certification of Central Air Conditioners - FINAL REPORT - APRIL 2003)

Ratios of total consumptions for EU 15 countries weighted all sectors and systems

(kWh/m².year) Cooling Heating Cooling & Heating

Austria 26,1 147,1 173,2 Belgium 19,5 135,6 155,0 Denmark 14,1 170,4 184,5 Finland 15,0 192,1 207,1 France 32,6 114,6 147,3

Germany 22,8 152,6 175,3 Greece 48,3 97,2 145,4 Ireland 19,5 119,4 139,0

Italy 50,1 94,5 144,7 Luxembourg 19,1 135,9 155,1 Netherlands 17,7 136,0 153,7

Portugal 49,7 96,0 145,7 Spain 81,5 28,5 110,0

Sweden 14,9 192,1 207,0 UK 19,7 119,5 139,2

Electricity consumption for cooling (system and auxiliaries) in offices buildings for CAV (Constant

Air Volume system) and RAC (room air conditioner)

AC consumption (kWh/m2.year) Seville-CAV 99,26 London-CAV 32,77 Milan-CAV 70,49

Seville-RAC 58,52 London-RAC 8,39 Milan-RAC 28,53

Some indicative ratios for air conditioning consumptions for UK different sectors

(All ratios are based on based on cooled area)

AC (kWh/m²) Central systems

AC (kWh/m²) Packaged systems

Ventilation (kWh/m²)

Hospitals (all health sector) 94 170 71

Hotels/restaurant/bar 94 170 71 Offices (commercial only) 49 270 117

Commercial establishments (retail) 144 123 52

Collective housing Schools (all education) 94 170 71

Leisure 94 170 71 Government buildings 94 170 71

Warehouses 94 170 71

17

2) Some indicative ratios for French offices from an internal study of the Ecole de Mines:

Use Additional remarks Ratios All consumptions The ratio take into account the

climate correction kWh/m².year

240

Heating and cooling The ratio take into account the climate correction

kWh/m².year

150 - 170

Cooling The ratio take into account the climate correction

kWh/m².year

25 -50

DHW kWh/m3 70 (electric) 100-150

Lighting kWh/m².year 40 - 65 Lift As % of total energy

consumption 2

18