1. Technology case description Lighting Task 21 Experts meeting, Seoul 19 April 2001 Harry Vreuls Operating Agent

2. 5 Case applications France Energy Savings Calculation for CFLs in the household sector 2 December 2010 Netherlands Lighting in households 1 March 2011 version 2 Korea Energy Savings Calculation for ballasts for 32W fluorescent lamps 10 February 2011 Spain Lighting. Household January 2011 USA Energy and Demand Savings Calculation for 2006-2008 Upstream Lighting Programs in California, 21 January 2011

3. Elements for overview (1) General – Program name and country (title) – Technology involved (1.3) – Status of the calculation (1.4) Formula – Formula used (2.1) – Parameters (2.2) – Baseline issues (2.3) – Normalisation (2.4) – Corrections (2.5)

4. Elements for overview (2) Energy Savings – Calculation of the annual savings (3.2) – Calculation of the lifetime savings (3.3.2) Greenhouse gas reductions) – Calculation of the annual GHG reductions (4.1.2) – Calculation of the lifetime GHG reductions (4.2.2) Remarks – Warnings for users, general opinions experts

5. General Technology involved (1.3) France: Compact fluorescent lamps (class A EU label) Korea: – Ballasts for 26mm 32W fluorescent lamps(FLR) – Ballasts for 16mm 32W fluorescent lamps(T-5) – Ballasts for 32W U-shape fluorescent lamps(FPL) Netherlands: CFL lamps Spain: The main types of lamps that can be used are: – CFL: Compact Fluorescent. – Metal halide lamps. – ED (Light Emitting Diode) USA: – Screw-in CFLs : basic bare spiral CFLs, as well as several types of specialty CFLs (e.g., dimmable, three-way wattage, reflector-style, A-lamp shaped, and globe-shaped). – Energy Efficient Lighting Fixtures hard-wired, compact fluorescent (CF) interior and exterior lighting fixtures; CF torchiere lighting fixtures, plug-in fluorescent desk, table and non-torchiere floor lamps – Various types of LED lighting products: LED nightlights, LED holiday lightstrings, LED open/close signs, LED desk/task lights

6. General Status of the calculation (1.4) France: – The evaluation are officially used to measure the energy savings related for different standardised operations and equipment within the white certificates scheme. They are published in French on the Ministry web site and in a print by ATEE (“Mémento du Club C2E). – A summary description of the detailed calculation sheets with all assumptions are restricted to the stakeholders involved in the process of fixing the standard savings values. Korea: – Electric power infrastructure center produce a performance report for internal use – Electric power infrastructure center updates the parameter used to calculate the energy savings through irregular survey, and the calculated energy savings is determined as business performance after center’s review of subsidy program result

7. General Status of the calculation (1.4) Netherlands: – This case application holds a way how to calculate the energy savings of lighting in households in an aggregated way. – In the program the savings are not reported in an aggregated way Spain: no information USA : – The evaluation approaches are varied but in general follow various guidelines including the IPMVP and the NAPEE Impact Evaluation Guide as well as state specific guidance such as the California Energy Efficiency Evaluation Protocols – Final Evaluation Report: Upstream Lighting Program Volume 1 CALMAC Study ID: CPU0015.01 February 8, 2010 Available at www.calmac.org

8. Formula: Formula used (2.1) – France: Annual savings in year t = (1- MSEFF) * {ns x [1/1000 x (Pold x hold – Pnew x hnew)] } in kWh – Korea: E saved = P saved x H annual x N s – Netherlands: Annual electricity savings in year t = n s x [1/1000 x (P old x b o – P new x b n )] – Spain: Savings (kWh) = H * S * N * (Po - P), where Po = Σni * Poi and P = Σni * Pi – USA: UES (kWh/year): IRp x HOUp x ΔWp/1000 UES (peak kW): IRp x CFp x ΔWp/1000

9. Formula: Parameters (2.2) France: – Pold = the average capacity in W of the (old) bulbs – Pnew = the capacity in W of the (new) CFL – bo = bn = burning hours – ns = number of CFL units sold in year t – 1/1000 = conversion factor from W to kW Korea: – E saved = Energy savings (kWh) – P saved = Power savings per unit (kW/unit) – H annual = annual running hours (h) – N s = the number of subsidized units (unit)

10. Formula: Parameters (2.2) Netherlands: – For the values of important parameters data from annual household inventories are used – Use of average capacity (Watt) and the average burning hours (estimated) – The energy savings are deemed (average) savings – Annual sales data are not available ; data are estimated based on household surveys in the year 2003 and 2008 Spain: – H is the number of households in Spain – S is the number of lamps per house substituted. – N is the annual number of hours of lighting usage, around 700 in Spain – Po is the installed power before the measure – ni the number of lamps of one specific kind and Pi the power (in kW) per lamp of that specific type – P is the installed power (in kQ) after the retrofit; determined as Po, using new lamp characteristics USA: – IRp = installation rate for IOU-discounted product p – HOUp = annual average hours of use for IOU-discounted product p – ΔWp = average displaced wattage for IOU-discounted product p – CFp = average percent on at peak for IOU-discounted product p

11. Formula: Baseline issues (2.3) France: – An average capacity of the 'old' incandescent lamp (Pold) of 80 W is assumed – The number of burning hours is assumed constant (800 hours per year) – The baseline used for the energy savings calculations is a market average: assumed that a CFL replaces in 70% of the case an incandescent lamp and in 30% of the cases a CFL – This baseline is incorporated in the gross to net calculation – The baseline is static; the discounting (during the lifetime calculation) is considered to somehow take into account a dynamic baseline. Korea: – Baseline for the magnetic ballast is a 40W fluorescent lamps and the sum of electric power for a fixture (a fluorescent lamp and a ballast) is assumed to be 50W – Baseline is a stock average and it does not change during the business period – Burning hours before and after the replacement would not change Netherlands: – a reference situation is used and not the (real) before situation: a conventional or incandescent bulb. – an additional unit would otherwise have been installed with a conventional bulb with comparable capacity – replacement of (broken) CFL units is negligible – immediate replacement and use of CFL, so no ‘spare units’

12. Formula: Baseline issues (2.3) Spain: – The baseline is the energy consumed before the application of the measure and is considered to be static; there are no factors of influence – The lighting levels as well as the location of lamps within the house are assumed to remain constant USA: – For residential CFLs, the evaluator calculated the average wattage of non-CFL equivalents by lamp shape and room type. The evaluator then averaged the room-type non-CFL wattages, weighting by the room-type distribution of CFLs of that shape. – For nonresidential CFLs, self-report data were collected onsite to estimate the wattage of pre-existing equipment. Pre-existing wattages were estimated using regression techniques for various post-retrofit wattage categories – The wattage of baseline fixtures was estimated per fixture categorie rebated through the program and was assumed to be the same for both residential and nonresidential applications. – Baseline wattage was estimated as the average wattage for the particular lamp shapes installed in particular room and/or fixture types. – There was no empirical evidence of decreasing replacement wattages over time for CFLs

13. Formula: Normalisation (2.4) – France: not applicable – Netherlands: not applicable – Korea: no normalisation has been applied for the energy savings – Spain: there is no normalization – USA: the savings were normalized to annual values for hours of operation and peak coincidence factor as indicated

14. Formula: Corrections (2.5) France: – Annual savings are corrected to account for the average reference market share of CFL – Net savings are equal to gross savings, as calculated above, multiplied by a coefficient reflecting the market share of the efficient appliance (MSEFF) = gross savings *(1- MSEFF) – It is assumed that on average a CFL replaces in 70% of the case an incandescent lamp and in 30% of the cases a CFL. Netherlands: – CFLs are only bought for replacement of existing lamps. – Two elements could be taken into consideration for the calculation from gross to net: additional lighting (e.g. garden lightning) longer burning hours for specific lighting (e.g. whole night burning outdoor lightning) – There is some research indicating these rebound effects, but due to lack of reliable national data this is not taken into account – As there are no actions included in this case application, other topics as free riders and double counting are not relevant.

15. Formula: Corrections (2.5) Korea: No corrections has been made for the energy savings Spain: there are no significant interactions to be considered; no need for normalization as there are no influence variables. USA: – Several corrections were applied to the calculation of gross energy savings. – Table 4 summarizes the evaluation elements and inputs – The corrections were for: Shipments versus sales Leakage Net versus gross savings

16. Discussion How to ‘correct’ the information: – Netherlands: parameters – Normalisation: repetition of baseline One common formula? Conclusion(s) for each topic Other points

17. Overview general formula Nether lands USAFranceSpainKorea Formula based on energy savings = Old lamps – new lamps (general for all types of lamps) XXXXX Unit is lampXXXXX Burning hours old lamp = new lamps XXXXX Factor Kilowatt to watt 1/1000XXXXX Savings related to CLFXXXOO Savings related to LEDOxOXO Savings related to FLOOOOX Factor Watt old./. new lamp4,5 = CLF O4,45 (80/18) = CLF 10 (40/4 ) = LED 1,77 (32/1 8) = TL Normalisation factor replacement old CLF by new CLF OX (in NTGR) XOO

18. Energy Savings Calculation of the annual savings (3.2) France: Annual net savings in year t – = (1- 0.3) * {ns x [1/1000 x (80 x 800 – 18 x 800)] } – = 0.7* ns x 49.6 – = ns x 34.72 kWh Korea E saved (Ballasts for one 32W lamp, Residential use) in 2009 business performance – = 0.018kW/unit x 2,771h x 117,778units = 5,874,531kWh Netherlands: Annual electricity savings in year t – = (5,678,000 /1000) x (3.5 x 12.4) x 482= 118,776,946 kWh/yr (about 118.8 GWh/yr). Spain: Assuming a house with 5 lamps of 40 Watts, substituted by LED with 4 Watt consumption, annual savings are: – (kWh) = H * S * 700 * 5 * (0,04 – 0,004) = H * S * 126 kWh/yr

19. Energy Savings Calculation of the annual savings (3.2) USA screw-in CFL (more information in separate report – Detailed calculations resulting in net kWh/y peak kW

20. Energy Savings Calculation of the annual savings (3.2) USA screw-in CFL (more information in separate report – Detailed calculations resulting in net kWh/y peak kW

21. Energy Savings Calculation of lifetime savings (3.3.2) France – The life time savings are not used for how long savings are accounted for, but for accounting the savings of the CFL promoted in year t – The life time savings are discounted (saving in kWh cumac) with a discount rate of 4%. This results in the value of 6.626 year for the discounted lifetime (LTdisc) for CFL (life time of 7.5 years) – Lifetime savings: LTdisc x ns x 34.72 = 6.626 x ns x 34.72 kWh Korea: E saved, lifetime (Ballasts for one 32W lamp, Residence) in 2009 business performance – = 7years x 0.018kW/unit x 2,771h/year x 117,778units = 41,121,718kWh Netherlands: – Lifetime savings=12 x 118,8 GWh = 1,425 TWH

22. Energy Savings Calculation of lifetime savings (3.3.2) Spain: The total lifetime savings are: – Lifetime savings: H * S * 126 * 71,4 = H * S * 9.000 kWh/yr USA – For this evaluation only annual demand and energy savings were calculated – For other evaluations, measure life times are estimated by dividing the annual hours of operation into the measure lifetime as provided by manufacturers or from databases such as DEER

23. Discussion For each case application one (or more?) an energy savings value? Why are the savings different? This is clear from: – Korea: FL (long annual burning hours) – France: CFL replacing CFL – Spain: LED as baseline – USA: detailed corrections – Other remarks? Lifetime savings based on technical lifetime divided by annual burning hours

24. Greenhouse gas reductions Calculation of annual reductions (4.1.2) France: There was no calculation of GHG savings Korea: – GHG emission factor of 2007 used: 445g/kWh – Annual GHG savings(CO2e) = 31,614MWh x 445g/kWh = 14,068,230kg = 14,068 ton Netherlands: – Using GHG emission factor (0,566 kg CO 2 /kWh) the annual GHG savings are estimated as: = 118,776,946 kWh x 0,566 kg CO 2 /kWh = 67,227,751 kg CO 2 (about 67.3 Gg CO 2 Spain – Using medium emission factor for the national electrical system: 0,360 kg of CO2 per kWh – Annual savings of CO2 are: GHG Savings (kg CO2) = H * S * 0,360 * N * (Po-P) – There are no other GHG affected significantly by this measure other than CO2. USA: There is no calculation of GHG savings

25. Greenhouse gas reductions Calculation of lifetime reduction (4.2.2) France: There was no calculation of GHG savings Korea: – GHG lifetime savings = lifetime x annual energy savings(kWh) x GHG emission factor of electric energy(g/kWh) – = 7 x 31,614MWh x 445g/kWh = 98,477,610kg = 98,477 ton Netherlands: – 118,776,946 kWh x 0,566 kg CO2/kWh x 12 = 806,733,017 kg CO2 (about 806.7 Gg CO2) Spain – GHG lifetime savings = lifetime x annual energy savings(kWh) x GHG emission factor of electric energy(g/kWh) – = 7 x 31,614MWh x 445g/kWh = 98,477,610kg = 98,477 ton

26. Discussion GHG information is in 2 cases missing; to be estimated? GHG emissions, only CO2 included Country specific factors – Korea: 0.445kg CO2/kWh – Netherlands: 0.566 kg CO 2 /kWh – Spain: 0.360 kg CO2 per kWh

27. Remarks Warnings for users, General opinions experts Others?

28. Approaches for all technologies – Overview for selected elements for each technology – Discussion on including missing information on the expert meeting – Discussion on one (or two) common formulas for a technology – Include these overview in the general report – Present the common formulas separate in the report