EPLabel REPORT ON PROPOSED ENERGY BENCHMARKING SYSTEMS FOR SIX SECTORS. By Robert Cohen ESD Project Co-ordinator. And

Transcription

1 EPLabel Technical Report presented by Energy for Sustainable Development Limited P00937 REPORT ON PROPOSED ENERGY BENCHMARKING SYSTEMS FOR SIX SECTORS By Robert Cohen ESD Project Co-ordinator And Bill Bordass, William Bordass Associates John Field, Target Energy Services Project Sub-contractors December 2006 E N E R G Y F O R S U S T A I N A B L E D E V E L O P M E N T

2 This document has been written within the framework of the SAVE EPLABEL project. This project is co-funded by the European Commission DG TREN, within the IEE framework. The sole responsibility for the content of this publication lies with the authors. It does not represent the opinion of the European Community. The European Commission is not responsible for any use that may be made of the information contained therein.

3 CONTENTS INTRODUCTION... 5 A GRADUATED RESPONSE METHOD OF APPROACH... 6 STATE OF ART REVIEW... 7 DEVELOPING THE BENCHMARKING SYSTEM... 8 BACKGROUND TO ENERGY PERFORMANCE BENCHMARKS FOR BUILDINGS IN USE... 8 EPLABEL S APPROACH TO BENCHMARKS... 9 IMPLEMENTING THE BENCHMARKING SYSTEM IMPLEMENTATION OF BENCHMARKING LEVELS 1 AND TESTING EPLABEL LEVEL 1 AND 2 BENCHMARKS IMPLEMENTATION OF LEVEL 3 BENCHMARKS FOR OPERATIONAL RATINGS CONCLUSIONS EPLabel interim technical implementation report December 2006

4 EPLabel interim technical implementation report December 2006

5 INTRODUCTION BACKGROUND The EPLabel project addresses the EU Energy Performance of Buildings Directive (EPBD) Article 7.3: the requirement for Public Buildings over 1,000 m² to display an Energy Certificate prominently 1. It started in January 2005 and finishes in early 2007 and involves nineteen countries, ten with full Partners 2. The project s main technical objective is to develop a methodology for energy benchmarking and certification based on Operational Ratings (actual annual energy consumption). It covers six sectors: Public administration offices, Universities, Schools, Sports facilities, Hospitals and Hotels. THE BENEFITS OF OPERATIONAL RATINGS In buildings with ongoing activities, a well-designed procedure of energy certification based on Operational Ratings has several important advantages: a) it can be easy, low cost and appropriate; b) it can build on existing energy management activities and initiatives; c) it can get started quickly, without the need to train experts; d) it can incentivise building owners, managers and occupiers to make immediate reductions in energy use and CO2 emissions; and to seek more advice where necessary; e) it is easy to understand by the public; and f) it can help to close the gaps between client and design aspirations and actual performance. The aim of the EPLabel project is to ensure that the implementation of Operational Ratings can capitalise on all these advantages. FIVE STEPS For ease of description, we have broken down the process required to produce a certificate based on an operational rating in five main steps. What have we got? STEP 1 Collect relevant data and calculate the appropriate building energy performance indicator(s). What does it mean? STEP 2 Identify appropriate benchmarks with which the indicators can be compared. STEP 3 Compare the indicators with the benchmarks to grade the building s energy efficiency. What can we do? STEP 4 Identify cost-effective energy saving measures. Tell everyone about it: STEP 5 Bring together all the relevant data onto the energy performance certificate (EPC). As far as the user is concerned, these steps will be combined in an integrated system with a simple entry level which can get the process going while satisfying the requirements of the EPBD. The EPLabel project has been developing robust approaches to each of the above five steps. Our specific focus has been on Step 2, identifying benchmarks, the subject of this report. 1 OJEC - Official Journal of the European Communities (4 Jan 2003). The Energy Performance of Buildings Directive: Directive 2002/91/EC of the European Parliament and of the Council of 16 December 2002 on the Energy Performance of Buildings. 2 The project Co-ordinator is the UK Partner, Energy for Sustainable Development Ltd, who are supported by sub-contractors Target Energy Services, William Bordass Associates and the Association for the Conservation of Energy. The project is funded in the UK by a Government Ministry (DCLG) and Constructing Excellence. The other partners are BBRI (Belgium), Energiereferat Frankfurt (Germany), Esbensen (Denmark), CSTB (France), NKUA (Greece), NUID (Ireland), DHV (Netherlands), Enerma (Sweden) and Motiva (Finland). EPLabel interim technical implementation report Page 5 December 2006

6 A GRADUATED RESPONSE METHOD OF APPROACH Each of the five steps needed to obtain an Energy Performance Certificate (EPC) based on an Operational Rating can be undertaken with different degrees of rigour. The EPLabel project is developing a graduated response which allows a robust yet pragmatic progressive introduction of EPBD Article 7.3 to suit the knowledge available in each country or region for each building sector or sub sector and the level of resources an organisation is able to apply 3. An easy entry level is proposed for cases where detailed information is hard to get or may be less rewarding, whilst more detailed assessment is suggested where the need and scope for improvement is greater, or where the owners, occupiers or management of the building are interested in a more thorough assessment. A cohesive framework makes assessments at different levels as consistent as possible. BUILDINGS LARGE AND SMALL By number, most non-domestic buildings are small in size and most are straightforward to understand and to assess - provided that key information is available. However, about half of the total nondomestic floor area is in the larger and more complex buildings (over about 1000 m2 each), although these only account for some 5% of premises by number. These larger buildings also include many of the more intensively-used and energy-intensive ones and hence account for more than half of nondomestic energy use. Larger buildings also tend to have dedicated management which potentially could improve energy performance substantially, even if, for the most part, this has not been a high priority in the past. ENERGY INTENSITY AND ENERGY EFFICIENCY With pressures to get more from less, organisations today are encouraged to intensify the use of their buildings, e.g. with longer hours of use (e.g. more community use of schools in evenings and at weekends) and higher levels of occupancy and use of equipment. Some buildings also contain energyintensive facilities (e.g. restaurants and server rooms) which are either not present in typical buildings of the same type, or are usually much more modest. Managers are concerned that an inappropriate system of energy certification might penalise these intensified buildings for what may actually be an increase in operational efficiency per person or unit of production. ACCOUNTING FOR HIGH-INTENSITY BUILDINGS In the past, high-intensity features have been dealt with by: a) Having different benchmarks, e.g. for schools with swimming pools or offices with restaurants and computer rooms. However, the special facilities often vary considerably in size and energy use, and so the allowances can easily be inappropriate, and can also be misused. b) Normalising energy consumption down for hours of use a feature which has often been abused by making over-generous corrections. We therefore suggest that the entry level of analysis permits few if any exceptions, so energy-intensive buildings get a relatively harsh grade, even if they are in fact quite efficient at what they do. Detailed analysis (motivated by the market and legally compliant, but not forming part of the statutory minimum requirement) could then put the callipers on such exceptions, for example permitting: a) allowances for energy use in, say, a computer room, only if the area had been sub-metered and had had an accredited energy survey, with energy-saving recommendations; b) allowances for high occupancy densities or long operating hours only if verified and signed off by a responsible person who would take the consequences if the allowances subsequently proved not to have been justified. PAYING FOR ADDITIONAL DETAIL Ideally, the driver to more complex analysis would come from the market, e.g. with providers and users demanding more detailed analysis of energy use in buildings which they consider to be more energy efficient than a simple grading shows. Motivators would include meeting performance targets, corporate responsibility, landlords demonstrating the energy-efficiency of their properties, tenants striking bargains, and owners, managers and occupiers (and their service providers including facilities and maintenance managers, managing agents, and client, design and building teams) all striving to look better in their organisational performance and environmental reporting. 3 For more details see: Usable Buildings Trust (June 2005), Onto the Radar: How energy performance certification and benchmarking might work for non-domestic buildings in operation, using actual energy consumption. Downloadable from EPLabel interim technical implementation report Page 6 December 2006

7 STATE OF ART REVIEW At the start of the project, the Coordinator designed a questionnaire to identify the state of the art in each participant country and the associated countries. For each country and the six building sectors in EPLabel, the questionnaire dealt with: Application of EPBD in the sector Relevant division into sub-sectors (building sub-types) Type of spaces in each building sub-type Possible members of an Industry Advisory Group Energy data and building size metrics by sector and methods for climate/weather normalisation Energy benchmarks availability Publications, case studies, etc The full results from this review have been written up in a separate deliverable. The main conclusion in respect of benchmarks is that the Partner countries can be divided into three groups: Benchmarks available for most sub-sectors Countries in this group are UK, Denmark, Germany and Finland Benchmarks available for a few sub-sectors Countries in this group are Belgium, Sweden, Ireland and Greece No benchmarks available Countries in this group are France and Netherlands However, the quality of the benchmarks is highly varied, with some being based on national statistics with sample sizes extending to thousands of buildings, some on detailed studies and analysis of a relatively small set of buildings and some on statistics from a relatively small sample of buildings. EPLabel interim technical implementation report Page 7 December 2006

8 DEVELOPING THE BENCHMARKING SYSTEM BACKGROUND TO ENERGY PERFORMANCE BENCHMARKS FOR BUILDINGS IN USE What are benchmarks for? Benchmarks support the drive to buildings which use the smallest amount of energy and create the smallest greenhouse gas emissions in relation to their purpose. There are two fundamental ways benchmarks can be used: 1. To identify if a building s energy performance is good, average or poor with respect to other buildings of its type. This is a robust indicator of whether the building should be prioritised for energy improvement action. For this purpose, empirical benchmarks derived from stock energy statistics are applicable. 2. To identify if a building s energy performance matches its potential and if not by how much it might be improved cost effectively. For this purpose, a realistic model of the building and its systems is theoretically more applicable 4. EMPIRICAL BENCHMARKS Empirical benchmarks fall into two categories: 1) The conventional type are obtained from bulk statistical data, 2) Less common but arguably more insightful are parameterised benchmarks, as in ECON 19 5, which set criterion (Best/Good Practice) and normative (Typical) standards for each energy end-use in the building. The values are derived from detailed studies of typically 20 to 100 buildings per sub-sector (ECON 19 has four) involving energy audits and sometimes sub-metering; against a background of less well-defined bulk data where available. Parameterised benchmarks can be used as robust fixed benchmarks in place of statistics: ECON 19 has been the de facto benchmarking standard for UK offices since Furthermore, they lend themselves to being customised to the activities and use of an individual building. However, this does not mean that customised benchmarks for heating and cooling energy need to reflect the form and fabric of the building; they deliberately are related purely to the area of each activity, its intensity and its hours of use. MODELLED BENCHMARKS In theory a model of a building s energy use should be able to imitate reality and enable a forensic examination of any inefficiencies in its operation and the scope for improving its fabric and services. Such models have been under development for three decades and modern computing power does make this prospect tantalisingly close. An Asset Rating assumes a standard external climate, that the building s activities have a standard use, only calculates the energy needed for HVAC, hot water and lighting; and uses standard assumptions for the performance of automatic controls and management, according to the controls that are present. However, especially for the more complicated existing buildings, the information required as input to even a basic Asset Rating model can be difficult and expensive to obtain 6. Nevertheless, once such a model is built, it is simple to compare the actual building with a notional building i.e. a customised reference having the same form as the actual building but with different U values or plant efficiencies, and produce a grade. It is a quantum jump to extend an Asset Rating model to calculate a benchmark for an Operational Rating, as here it must be able to replicate to a satisfactory tolerance the measured energy performance of a building. This validated model must take account of the actual use of the building, the actual indoor environment, the actual external climate, the actual operation and control and represent all energy end uses. If achievable, such a model might be used for benchmarking and to predict the benefits of a package of operational and investment energy saving measures, in an approach akin to that currently recommended by the relevant draft CEN Standard 7. However, for getting a system of Operational Rating started, this is far too big a step to be practically possible. 4 A rapid, initial benchmark for this purpose would be the energy performance from, say, the two previous years of operation 5 Energy Consumption Guide 19 ( ECON 19 ), Energy use in offices, Action Energy (2003). Downloadable from 6 Often up to date plans and sections of the building and drawings of the mechanical and electrical plant may not exist. 7 The validated calculation model is a recommended but not mandatory procedure to assess the energy effectiveness of possible improvement measures in CEN pren (2005). EPLabel interim technical implementation report Page 8 December 2006

9 EPLABEL S APPROACH TO BENCHMARKS Central to the graduated response proposed by EPLabel is a unified scheme of empirical benchmarks for Operational Ratings at three progressive levels, as illustrated in Figure 1: Level 1 Simple (e.g. Derived from stock statistics or analysis) Level 2 Corrected (for special energy uses not included in the simple benchmarks) Level 3 Customised (to take closer account of what the building does). The Level 1 and 2 benchmarks may well be sufficient for smaller, simpler and more standard buildings and be adopted in initial statutory implementations of Article 7.3. For more complex buildings, the Level 3 benchmarking approach would allow more meaningful and fairer assessments of a building s energy use and CO 2 emissions. However, Level 3 benchmarking may be difficult to integrate into a statutory repeatable procedure. It may therefore first be introduced by sector or sub-sector as a voluntary good practice procedure, for expert use, where it can provide greater insights which will assist the dialogue between the investigators and the building owner; and can be used to inform and improve the less flexible (but more reproducible) procedures that governments are likely to require and which can be enforced and verified more readily at Level 2. Good Practice Typical ACTUAL Gas Electricity 77 kgco 2 /m kgco 2 /m kgco 2 /m Level 1: simple, usually reconciled with stock statistics for the type of building concerned. Good Practice Typical ACTUAL Gas Electricity 77 kgco 2 /m 2 Special (20 kgco 2 /m 2 ) 183 kgco 2 /m kgco 2 /m kgco 2 /m Level 2: corrected, taking account of special energy uses not included in the Level 1 benchmarks. The above charts show a building s actual carbon dioxide emissions in comparison with fixed Typical and Good Practice benchmarks appropriate to the building. With the level 2 approach, the emissions from a special energy use (not included in the benchmark reference) are identified by measurement (e.g. survey or sub-metering) and deducted from the total emissions before making the benchmark comparison. Level 3: customised, taking more detailed account of the building s schedule of accommodation, activities and use. The chart on the right includes benchmarks built up from each of the energy end uses occurring in the different activity areas in the building concerned, calculated separately assuming Typical and Good Practice energy efficiency for the fabric, plant, management and controls involved for the actual occupancy of the building (hours and densities of use). Tailored benchmark Typical performance ACTUAL ENERGY USE Tailored benchmark Good Practice performance Figure 1 Three levels of benchmarking for a measured Operational Rating kg of carbon dioxide per m² per year Gas heating and hot water Refrigeration and heat rejection Fans and pumps Lighting Electric humidification Office equipment Other normal, e.g. lifts Other special, e.g. computer rooms The component-based values used at Level 3 apply to iconic descriptions of stereotypical buildings in a sector: the building s schedule of accommodation (often simplified to principal activity areas, e.g. cellular offices, open-plan offices, restaurants), typical levels of occupancy and equipment and the hours of use. The principal parameters in this method can be summarised using the energy tree diagram (Figure 2), which goes down to the roots of consumption separating out the asset (standards and efficiencies), and operational (use, control and management) elements of energy use 8. Benchmark values can then be reported not just for the building as a whole, but at any scale - engineering system, 8 Field, J. (2006). Energy assessment and reporting method CIBSE TM22: 2006, CIBSE Publications, London. EPLabel interim technical implementation report Page 9 December 2006

10 activity area, zone, or smaller, and also for each energy end-use (e.g. lighting), and their components 9. This allows benchmarks to be re-computed if necessary to suit genuine differences, e.g. a building with a different set of activity areas or energy end-uses; or a higher intensity of use. The tree diagram description also suits what-if calculations, e.g. estimating the likely savings if the installed power density of the lighting was halved. A Total energy use per m 2 (primary or CO2 equivalent) B B Lighting kwh/m 2 Ventilation kwh/m 2 Other uses C Lighting W/m 2 D Effective hours/yr C Ventilation W/m 2 D Effective hours/yr E F G H Light level Efficiency Hours Management Lux (W/m 2 )/100lx of use factor E F G H Vent rate Efficiency Hours Management (l/s)/ m 2 W/(l/s) of use factor Asset Use, Control Asset Use, Control and Management and Management Figure 2 The Tree Diagram description of a building s annual energy use The example in Figure 3 illustrates the components of the published annual electric lighting energy use benchmarks for a standard air-conditioned office in the UK (ECON 19). Typical and Good Practice criteria are provided not just for the annual electricity consumption, but for each contributory element i.e. the light level, the efficiency of the lamps and luminaires, the hours of use and the control and management factor (which includes assumptions about the availability of daylight). We can combine together similar analysis for each end use in each area of this typical building to produce total energy benchmarks assuming performance at Typical or Good Practice level. Lighting annual energy use kwh/m² 27 Key: 54 Black is Typical Green is Good Practice X Installed load W/m² Effective hours/yr Light level 4.0 Efficiency 3.0 X x100 lux 5.0 (W/m²)/100lux 4.0 Occupied 3200 Control 70% X hours/yr 3200 factor 85% Figure 3 Using the tree diagram to summarise the benchmarks for lighting in an office Bringing statistical and parametric benchmarks together Parametric benchmarks allow one to get going on a useful system without having all the statistics, whilst statistical benchmarking is practical where good quality bulk data are available, but parametric components have yet to be defined. As mandatory energy certification becomes required across so many sectors; and where there are large variations in knowledge about energy performance (both by building type and by country), we need to combine the insight achieved by using the parametric approach and the practicalities of the statistical one. In fact, the two approaches can complement each other: statistical and parametric benchmarks can be directly linked at the median. One can associate and then reconcile the tree diagram breakdown of typical energy consumption for the stereotypical building with the energy performance of the stock median. Ideally the stock statistics will either exclude buildings with special substantial energy uses (e.g. a swimming pool at a school), or the sample size will be sufficiently large and the incidence of these specials relatively low, so that their influence on the median will be negligible It can also be applied at different scales for different systems and different boundaries simultaneously, whilst avoiding double counting of course. This can be useful when it is difficult to define unique boundaries. 10 Ideally, stock statistics would also exclude "special buildings" resulting from high hours and intensity of use; but again these should not affect the median if the majority of buildings in the sample are lightly-stressed. Perhaps a bigger problem is the "pool" one is fishing in - often one can get information most easily for the buildings with more active and interested managers. Efficient or not, these are often high energy buildings owing to higher intensities of activity and equipment, (e.g. big supermarkets) than in other buildings of ostensibly the same type. EPLabel interim technical implementation report Page 10 December 2006

11 IMPLEMENTING THE BENCHMARKING SYSTEM IMPLEMENTATION OF BENCHMARKING LEVELS 1 AND 2 CEN Draft Standard pren proposes two benchmarks are used: Rr: the value typical of the requirements of energy performance regulations for new buildings. Rs: the median value for the building stock. Rr is most important for new construction and Rs for existing buildings. The standards recognise that good benchmarking data may not be available: if Rs is not known, they recommend obtaining rough estimates by collecting data from a small subset. They also appreciate that Rr is difficult to define for buildings in operation. Therefore they say alternative definitions can be adopted for Rr until sufficient data become available on the operational performance of new buildings. The WP2 review revealed that there is substantial variation in the availability of energy benchmarks in the countries represented in EPLabel. Only the UK, Denmark, Germany and Finland have anything approaching a complete set. In other countries, data are available for certain sectors but are not considered definitive statistics from which official values for medians or quartiles might be generated. In principle, for the UK building stock, the Typical values from energy consumption guides published in the UK should be appropriate for Rs. In practice, however, a review in 2005 by the EPLabel team showed that UK benchmarks were in need of an overhaul. EPLabel has adopted a pragmatic approach to overcome these shortcomings. The implementation of the Initial benchmarking approach requires two benchmark values, Rr (regulatory standard) and Rs (stock median), for both the electricity and fossil fuel use, to be designated for each building sub-type in each country. The choice of building categories is thus strongly connected to the availability of corresponding benchmarks. Partners have been given three choices: 1. Use national building sub-types which have associated national benchmarks. 2. Use another country s national sub-types and their associated national benchmarks. 3. Use national sub-types without benchmarks. Where this is chosen, it will be possible to produce only a Level 0 Certificate i.e. just the Energy Performance Indicator, without an A to G grade. Due to their lack of national benchmarks, some countries are currently using sub-types and the corresponding benchmarks from other countries. To deal with the issue of setting values for Rr, many Partners have elected to set Rr at a fixed percentage of the Rs value eg 50% or 70%. This followed the lead of the UK where it was felt that the existing Good Practice benchmarks were even less robust than the Typical values and a fixed percentage across all sub-sectors could end a lot of fairly futile argument. CLIMATE AND WEATHER CORRECTION Most benchmarks are national values and, by inference, are related to an average national climate. Historically, in many countries, it has been customary simply to adjust the actual energy to take account of the difference between the local and this average national climate. This has the benefit of enabling all buildings in the country to be compared on a single basis. However, it suffers from two serious flaws: 1. In most buildings the energy used solely for space heating purposes is not separately metered. Where, for example, gas is used for heating, it is often also used for producing hot water and often also for cooking. In some cases, it will be used for co-generation (CHP) and/or absorption cooling. In all-electric buildings, identifying the electricity used for space heating may be even more problematic. Given that most countries have substantial regional climatic variations, the potential error that might be introduced by correcting an estimated value for heating energy use is unacceptably large. 2. By adjusting actual energy to a national average climate, one is effectively moving the building from its real location to an average location for that country. If one uses this adjusted building energy performance to assess the viability of energy saving measures, the results could be grossly misleading. The alternative of offering the adjusted performance to judge the building s energy efficiency but using the unadjusted performance to judge energy saving measures would be too confusing. We have also found it dangerous to adjust actual consumption, as touch is lost with reality; and adjusted consumption can be confused with actual consumption. In order to overcome these shortcomings, EPLabel is adopting a two stage process (see diagram): EPLabel interim technical implementation report Page 11 December 2006

12 1. Climate correction of benchmarks: In order to cope with regional climatic variations, EPLabel proposes adjusting the national benchmarks for fuel/thermal and electrical energy use in order to create benchmarks applicable to the regional average climate for the building. This is done by defining national and regional climate indices (eg degree days to a given base) for heating and cooling. The part of the benchmarks for fuel/thermal and electrical energy use which are heating and cooling climate dependent are defined separately for each sub-sector. This part is adjusted pro rata the ratio of the national and the building s regional climate indices. 2. Weather correction of actual energy: This will take account of the differences in weather experienced by the building during the year of the assessment compared with the regional average climate. It will also deal with local year-on-year weather variations experienced by the building, which is important when data for more than one year is being reported. Due to the inevitable approximations involved, EPLabel recommends that an adjustment of the actual energy for the regional average climate should only be made when year-on-year comparisons are being undertaken. To do the weather correction, the user enters the heating and cooling (if applicable) climate indices for the year of assessment for the building s region. The actual energy is climate corrected pro rata these climate indices and the 20 year average climate index for the region. We do not recommend making a weather correction for cooling unless the annual cooling energy has been sub-metered (to check it is consistent with the assumed part of the benchmarks for fuel/thermal and electrical energy use which is cooling climate dependent in that sub-sector). Metered energy Step 2: Adjustment from local weather to Regional climate Metered energy adjusted to climate of Regional average National benchmark adjusted to climate of Regional average Step 1: Adjustment from national climate to Regional climate National benchmark Diagram illustrating methodology for climate and weather corrections SPECIALS The EPLabel software identifies which energy end uses are taken into account in the benchmarks for each building sub-sector. This allows a user to specify and quantify any special energy use which is not already given an allowance. The procedure asks for a description of any special energy use - up to four different special energy uses can be separately entered. Verification is sought that the quantity of energy used by the special has either been sub-metered or signed off by an accredited expert (thus providing a verifiable audit trail). The user can then enter annual energy use for the special for any of the energy carriers for which energy use has been entered for that building. For statutory purposes, it will not be enough just to deduct the special energy use. For a deduction to be permitted, it will also be necessary to provide robust data (e.g. from sub-metering or from a standard, verifiable, estimation procedure) on the magnitude of the deduction, together possibly with information on the energy performance of the special feature deducted and the scope for improvement. When a verifiable special has been entered, its energy use is deducted from the actual energy use before the headline energy performance grade is calculated and shown on the energy certificate. However, the energy certificate also shows the total weighted energy of the building before specials have been deducted. On the certificate annex, the building s energy performance grade before specials are taken into account (ie before they are deducted) is reported. MIXED USE The EPLabel procedure (in Excel) allows a building or site to comprise up to four different building subsectors. In the on-line version the number of sub-sectors in a mixed-use building or site will be unlimited. The mixed use concept is that the benchmarks (electrical energy and non-electrical ie fuel/thermal) for the complete building are calculated as a weighted composite of the benchmarks for its component sub-sector parts, pro rata their share of the building s or site s total internal floor area. Many EPLabel interim technical implementation report Page 12 December 2006

13 mixed use sites have only one set of energy meters, so this approach allows the physical boundary of the building or site being assessed and the metered (measured) energy boundary to coincide, as required by the CEN Standards. If these do not coincide (e.g. for a building on a school site where there is only one set of meters) the boundary should be moved outwards - made larger - until the building (or site) and the metered energy boundary do coincide. In this case composite (pro-rata) benchmarks may be required to create a benchmark value for the mix of accommodation on the site, but these are generally simple to produce. Sometimes mixed use buildings are divided into virtually self-contained premises as far as building services are concerned, e.g. with independently-serviced shop units (which can be considered separately), social/recreational facilities (often ditto) an office section, and self-contained apartments. If the services are not fully subdivided, then one has the option of: Composite benchmarks for the ensemble. Landlord's Energy and CO2 emissions Statements for the landlord's services which then allow certification of individual premises to take place where necessary (see Interim Technical Report Annex A (Step 1). TESTING EPLABEL LEVEL 1 AND 2 BENCHMARKS PRELIMINARY APPRAISAL The first version of the EPLabel software (V1.2) incorporating Level 1 and 2 benchmarking was issued to Partners for testing at the end of March The primary objectives of this issue were as follows: To obtain a definitive set of building sectors and sub-types for each country To obtain corresponding benchmark data To confirm climate correction procedures To obtain idiomatic translations of all terms used in the software so it works in eight languages. All partners returned (most of) the necessary information and the UK team incorporated this into a revised version of the software which was sent to Partners at the end of May, just in advance of the Partners meeting at the start of June INITIAL TESTING Each partner reported on their experience of testing the software in their Primary Test Group (see chart below). Their reports were included in the Interim Technical Report Annex B. 1. British 2. Central 3. Nordic 4. SW 5. SE 6. Benelux 7. Baltic Group EU Isles Europe countries Europe Europe countries states Regional Group leader UK Germany Denmark France Greece Netherlands Sweden method 2nd Partner in Group Ireland Belgium Finland compliant Austria Denmark Spain with Subcontractors Poland, Norway CEN Czech Rep Standard Associated Switzerland Italy Netherlands Estonia, Organisation Lithouania Public Sector Offices Higher education Schools Sports facilities Hospitals Hotels & restaurants Primary test group Development Initial tests Secondary test group IMPLEMENTATION OF LEVEL 3 BENCHMARKS FOR OPERATIONAL RATINGS A customised benchmarking approach for sports centres has been implemented in EPLabel software Version 2.0 by the UK team and was also sent to Partners at the end of May, just in advance of the Partners meeting at the start of June. It derives directly from the procedure developed to benchmark the energy use in UK sports facilities 11. The principles of the process are as follows: 11 Field, J. and Bordass, W. (2001). Energy Consumption Guide 78 ( ECON 78 ), Energy use in sports and recreation buildings. Downloadable from EPLabel interim technical implementation report Page 13 December 2006

14 1. A building is divided into a schedule of accommodation which comprises a set of up to about ten different activity areas (see examples in Figure 4). A default schedule for each building sub-type provides the user with a starting point for customising the areas to the building being assessed. 2. Each activity area is associated with Good Practice and Typical parametric benchmarks per m² for all the energy end uses needed in that area, with electricity and heating fuel treated separately. 3. Additionally, special energy uses (not included in the parametric benchmarks) can be added. Often these also have Good Practice and Typical benchmarks per unit of their size. 4. The software then simply sums the electricity and heating fuel identified for each end use in each activity area to produce the building s total electricity and fossil fuel (heating) benchmarks. 5. The customised benchmarks are shown graphically, broken down by either energy end use or activity area (see example in Figure 5). The parametric benchmarks (in step 2 above) for each energy end-use (e.g. lighting) in each activity area are built up using an energy tree diagram, which goes down to the roots of consumption separating out the asset (standards and efficiencies), and operational (use, control and management) elements of energy use, as described above. Benchmark values can then be reported not just for each energy end use in each activity area, but also, for their components e.g. lighting W/m²/100 lux, ventilation W/l/s, small power W/workstation or W/m², etc. The method, in principle, also allows the energy benchmarks for each activity area to be customised for the intensity and hours of use, where these can be demonstrated to deviate from standard values. Figure 4 Illustrative graphics of the activity areas in two types of sports facility. System energy benchmarks Activity zone energy benchmarks Good Average Special Other Pumps Ventilation Lighting Hot water Heating Good Average Features Common Changing Studio/courts Other sports Pool hall Figure 5 Customised benchmarks (in kg CO 2 /m²/yr) split by energy end use and activity area. EPLabel interim technical implementation report Page 14 December 2006

15 CONCLUSIONS BACKGROUND Two of the key deliverables of the project are a web based energy certification tool and demonstrator software illustrating how an energy certificate based on an Operational Rating might be produced. The precise nature of these deliverables was not clear at the proposal stage and indeed during at least the first year of the project because of the uncertainty created by delays in Member States announcing how they planned to implement EPBD Article 7.3. It had been anticipated that national legislation would be in place by 4 January 2006, the deadline for transposition in Article 15 of the Directive. In fact, in most countries, legislation is still not published, although during 2006 progress has been seen in the number of countries making pronouncements of their intentions. Additionally, the timetable in the Contract anticipated that the CEN standards would be nearing completion when EPLabel started. In reality the draft standards are still being finalised, creating further uncertainty as we must aim to make our procedures compliant with these standards. CONTEXT The EPLabel project does not have the luxury of waiting any longer for official plans to be announced and the CEN Standards to be finalised. At the Partners meeting in Greece at the beginning of June 2006, decisions had to be agreed on how to complete the project in a way that could provide the most benefits for Member States. The conclusions came as a result of the following factors: The graduated response to benchmarking has considerable attractions for many Member States. However, to avoid the customised (Level 3) benchmarking at the heart of the EPLabel proposal being sidelined as too advanced, it is imperative to offer Member States an entry level (Level 0) for situations where benchmarks do not yet exist and the Initial benchmarking (Levels 1 and 2) approach as options for their preliminary implementations. There is considerable work to be done in developing robust procedures at Levels 0-2. There is a hugely important task in harmonising such procedures across the EU, another key aim in the EPLabel proposal. It is essential that any proposed customised benchmarking approach is seamlessly integrated with the initial benchmarking concept. Level 3 benchmarking may be difficult to integrate into a statutory repeatable procedure. It may therefore first be introduced as a voluntary good practice procedure, for use by experts as a diagnostic tool to provide greater insights. INTEGRATED SOLUTION Our proposed solution to all these requirements is to produce a mutually consistent triumvirate of the same core product: 1. A detailed written specification for producing an Energy Performance Certificate based on an Operational Rating and using the Initial benchmarking approach (Levels 0, 1 and 2). The EPLabel UK team has produced a preliminary vision of such a specification for the UK government. Because it is derived from the generic principles developed by the EPLabel project, including the graduated response and harmonising framework, it is, we believe, applicable across Europe. 2. Demonstration software comprising an Excel version of an implementation of the specification, aimed primarily at experts. 3. A web based version of the Excel software, aimed primarily at building and energy managers. Additionally, we propose to produce an advanced version of the demonstration software (in Excel) which will illustrate an integrated Level 3 customised benchmarking approach. BENCHMARKS The main outputs of this work will be an approach to benchmarks expressed in delivered energy terms for both electricity and fossil/thermal energy; as these continue to be the primary information on which effective benchmarking systems rely. These delivered energy benchmarks generally need to be used in conjunction with associated data (such as weighting factors for each fuel for CO2, primary energy etc. to suit national priorities) and assessment techniques (such as taking account of renewable energy supplies) which are core aspects of the Operational Rating process. The use of energy benchmarks can be simple or sophisticated. We identify three linked levels: EPLabel interim technical implementation report Page 15 December 2006

16 1. A simple legislative level using electricity and fossil/thermal energy benchmarks scaled by weighting factors and then combined to allow comparison with total headline weighted emissions per m 2. This level can generally be assessed by a non-specialist and is expected to cover a large percentage of all assessments - maybe 75% to 95% of all non-domestic buildings by number. For this purpose, we are hoping to develop simplified benchmarks for ordinary buildings without special activities and end uses. 2. Permitted corrections to the legislative benchmark comparison for use where the simple benchmark does not include a particular activity or energy end-use (e.g. a large data processing centre in an office) and so may not give a fair result. To make such corrections would require specialist and accredited assessors, who would need to use robust verification procedures: these would usually require additional measured data (e.g. sub-metering of the special end use) and comments on its energy efficiency. 3. A non-legislative diagnostic and advisory level which can be used to compare buildings with their peers, to set priorities, to help identify where the energy is being used and where problems exist (for example which services have poor energy performance - and possibly why) and to define likely areas for improvement, possibly with detailed targets. This activity will often require advice from specialists who are able make valuable judgements based on professional skills and often sparse data. This advice will therefore not be reproducibly verifiable and is hence not appropriate for legislative purposes. However, it will often be highly effective in identifying savings measures which can be put into effect rapidly and cost-effectively. Other issues involved with benchmarking have been addressed: A framework has been designed which helps to resolve technical issues such as differing floor area definitions and climate/weather correction. Benchmark values have been collated across similar building categories with the aim of eliminating contradictions which may cause problems in a legislative operational rating scheme. Opportunities for simplification of the benchmark sets have been identified in principle. Traditional benchmarks based on simple statistics for the median have been integrated with parametric benchmarking schemes (based on service levels, efficiencies and power densities) which provide far greater diagnostic potential. SUMMARY The EPLabel project has developed a framework for Operational Ratings to be implemented in a harmonised way across the EU, particularly as part of legislative procedures for EPBD Article 7.3 which requires larger public buildings to display an energy certificate to the public. We have developed a robust simplified approach to a customised benchmarking system which includes: Regionalisation of national benchmarks to allow for climate differences Mixed-use (composite) benchmarks for buildings and/or sites made up of different building subsectors Adjustments for special energy uses. We will also demonstrate a procedure to produce the more sophisticated customised benchmarks based on individual activity areas and intensity of use. However, because it was very difficult to get the simple levels to work across all countries, we have spent massively more time than anticipated achieving engagement. We will not be producing user versions of the customised benchmarks approach because we do not consider it suitable for an initial statutory implementation of Operational Ratings and will only distract everybody from getting Levels 0, 1 and 2 working smoothly. We hope and believe that the on-line benchmarking procedure we will be launching by the end of January 2007 will provide precisely what Member States will be looking for from the EPLabel project. EPLabel interim technical implementation report Page 16 December 2006

17

18 Energy for Sustainable Development Limited Overmoor, Neston, Corsham, Wiltshire SN13 9TZ United Kingdom Tel: +44 (0) Fax: +44 (0) Website: