1. 1 Summary of Reviews: Workpapers Approved by the California Technical Forum Meeting: California Technical Forum December 3, 2015 Jeff Hirsch/Kevin Madison Consultants - CPUC Ex Ante Team

2. Workpapers Reviewed by EAR Team Commercial variable refrigerant flow (VRF) Residential HVAC quality installation (QI) Clothes washer recycling High efficiency clothes dryers Advanced “Tier 2” power strips (two workpapers: residential and commercial) Condensing unit heaters Circulating block heaters for standby generators LED ambient fixtures and retrofit kits Retail Plugload Platform (RPP) including five measures (Freezers, Room air conditioners, Room air cleaners, Clothes dryers, Entertainment sound bars) 2

3. Variable Refrigerant Flow Challenges for the existing as well as proposed update to the deemed offering compared to standard deemed approach: –Proposed baseline and measure cases use different design and configuration normally only accepted for custom projects –Fuel switching from gas to electric heating assumed which invokes a policy based threshold analysis step –Calculation and modeling methods are relatively new, still under development, and thus require more validation of estimation long term accuracy 3

4. Source of Estimated Savings Majority of savings come from the difference in air distribution system design and operation rather than component efficiency improvements (from 8/4/2014 PA workpaper development presentation) 4

5. Deemed Measures are Typically “Like for Like” All other HVAC (packaged and split ac/ and HP, chillers, boilers, etc.) estimates are based on same configuration, capacity and operation, but higher efficiency Lighting based on same overall output, but lower input wattage Refrigerators have the same features (volume, freezer location, etc.), but lower annual kWh rating 5

6. Technology “Shifts” Do Not Change Nature of Service Magnetic bearing, multi-compressor centrifugal chiller uses a conventional centrifugal chiller baseline LED fixture uses either LF or HID baseline with similar workplane lighting level Instantaneous water heater uses storage water heater baseline with similar delivery capacity 6

7. EAR Team Concerns Is the assumed installed VRF measure system configuration typical? For NC, ROB, normal replacement and major renovation applications, are the baseline system assumptions typical of standard practice? Is the assumed program influence over the design decision plausible? 7

8. EAR Team Direction 1.Collect information on historical participant installations – building type, system configuration, operation, etc. 2.Determine the nature of project – NC, NR, etc. 3.For retrofits determine the pre-existing equipment configuration and operation 4.Understand the underlying design considerations that led to the selection of the VRF system choice and its configuration. How does EE program delivery fit into VRF selection. 8

9. Fuel Switching Policy “Fuel substitution programs may offer resource value and environmental benefits. Fuel- substitution programs should reduce the need for supply without degrading environmental quality. Fuel-substitution programs, whether applied to retrofit or new construction applications, must pass the following three-prong test to be considered further for funding: 1.The program must not increase source-BTU consumption. Proponents of fuel substitution programs should calculate the source-BTU impacts using the current CEC-established heat rate. 2.The program must have TRC and PAC benefit-cost ratio of 1.0 or greater. The TRC and PAC tests used for this purpose should be developed in a manner consistent with these Rules. 3.The program must not adversely impact the environment. To quantify this impact, respondents should compare the environmental costs with and without the program using the most recently adopted values for residual emissions Additionally, fuel substitution proponents must compare the technologies offered by their program with the most efficient same-fuel substitute technologies available to prospective participants that would have TRC and PAC benefit-cost ratio of 1.0 or greater.” Source: D.05-04-051, Attachment 3 (Policy Manual Version 3), page 10, Cost-Effectiveness Rule IV.10 9

10. Fuel Switching Considerations Arises when baseline versus measure have different energy sources (electric versus gas) Analysis of historical participants can inform the determination of typical baseline For the VRF workpaper, the measure includes a change in air distribution system, not just the VRF technology Baseline system should consider other viable high efficiency distribution system choices available to program participants 10

11. Modeling Methods and Applications 1.Chosen simulation tool must produce reasonable and comparable estimates for the measure and baseline systems 2.Analysis approach and equipment plus operation parameter assumptions must be consistent and in agreement with DEER (NOTE: not identical) 3.Detailed sample results analysis should show that savings estimates reflect real differences and not modeling anomalies 11

12. Simulation Tool: EnergyPlus 1.Does EnergyPlus (or any tool chosen to develop savings estimates) model all technology features that may significantly impact results (increase or decrease effective efficiency)? 2.Are these effects observable (and observed) in the results? Are the observed estimated savings due to the baseline and measure differences or other simulation tool anomalies or issues? 12

13. Simulation Tool: EnergyPlus VRF model examples: 1.Pipe length and relationship to (terminal unit and condensing unit) capacity and efficiency 2.Operating conditions; interactions with equipment components and controls; relationship to outdoor unit controls; resultant capacity and efficiency indoor unit entering wetbulb and outdoor unit entering conditions – versus suction temperature controls simultaneous heating/cooling Tool issues examples: 1.Plenum modeling (cannot currently use – known limitation) 2.Duct losses (explicit modeling – rather than T24 efficiency adjustment - requires use of airflow network) 13

14. Moving Forward: Approach and Assumptions Required Demonstrate that methods yield similar savings results to DEER estimates or adequately explain how differences are appropriate Not Required “Copying” of all DEER inputs Example of Possible Approach Show that proposed approach yields similar savings for conventional DEER measures such as a SEER 18 package AC and HP system upgrades 14

15. Moving Forward: Results Comparison Background Current DEER and workpaper methods compare ‘like’ or very similar technologies: simulation anomalies tend to error in “the same direction” and cancel each other. Concern When comparing fundamentally different technologies, if anomalies for each technology error in “opposite direction” then they either add to, or subtract from, each other. 15

16. Insights and Observations: Other Workpapers Use of DEER values and methods Standard practice baselines Best available data Coordination and timing with CPUC staff and ex ante consultants 16

17. DEER Values Use of DEER values and/or assumptions is generally required. Example DEER NTG values are to be used as default except when supported by more recent evaluation with staff approval Room AC and Freezer measures are covered by DEER DEER defaults used until recent participant results are available and approved by staff 17

18. DEER Methods Use of DEER methods is required, but methods as used here does not imply simply adopting DEER point values. Examples 1.Code baseline calculations for LED fixtures and retrofit kits - workpaper has an acceptance approach, but took time for EAR team review that could be shortened. 2.Demand impacts and interactive effects for clothes washers and dryers - incorrect demand analysis and incorrect application of HVAC IE to all loads could have been easily corrected if proposed approach had been outlined and reviewed at the abstract (or other) early stage. 18

19. Standard Practice and Code Baselines Cases when gross savings shall be calculated over the standard practice or code baseline: –ROB/NC/CE savings are above ISP/Code for full EUL –ER savings are above ISP/Code for post RUL period In general, an “internal consensus” (CalTF, PA or implementer) on standard practice is not adequate, especially if evidence indicates a higher efficiency standard practice. 19

20. Standard Practice Baseline: Clothes Dryers Given the high cost of the electric HP dryers, ex ante team views it unlikely that participant clothes washers will be just minimally compliant. –Clothes washers are likely more energy efficient, with lower remaining moisture content, thus reducing dryer energy use. –What available research might provide this information? Did the Pacific Northwest research look at the installed washers as well as dryers? 20

21. Standard Practice Baseline: Clothes Washer Recycling 1.Program stated motivation is that 50% of all discarded appliances are transferred to new owners yet analysis assumes that all program collected appliances would have been transferred to a new owner. –It is unclear how collecting the 50% otherwise destroyed (or even collecting broken machine) can be avoided –Savings values must be adjusted down to consider the fraction of units that would have been destroyed, as standard practice, without the program. 2.Standard practice baseline should assume all clothes continue to be washed in washing machines –Workpaper assumes 50% of collected machine get 100% savings thus assume those clothes are washed by hand. 21

22. Standard Practice Baseline: Residential HVAC QI CalTF “consensus” was that non-permitted system replacements rarely, if ever, include additional treatments such as duct sealing or air-flow adjustment. –This leads to the use of “test-in” results from SCE QI program data as the standard practice baseline This is not supported by WO32 non-participant sample, where 60% of sites had non-permitted HVAC replacements. –Observed system characteristics do not support use of the SCE “test-in” results as the baseline –Observed results indicate that contractor action to address other system faults or issues is the likely appropriate baseline 22

23. Best Available Data: Not Always Adequate Examples 1.Commercial advanced powerstrips: –The available field research, limited to a small group of university and college buildings, is not likely applicable to most other building types –PAs not likely to pursue a program for universities and colleges only 2.RPP soundbar measure –Operating hours taken from PG&E sponsored Nielsen research on television viewing hours which do not take into account any audio only usage –May overestimate typical hours of standby mode and the savings estimates are derived from standby hours 23

24. Best Available Data: Not Always Comparable – Part 1 Example For clothes washer recycling workpaper, measure and base energy use a mix of non-comparable values –Baseline (collected appliance): “Non-Energy Star” appliances monitored in 2006-2008 evaluation –Measure (counterfactual appliance): From DOE technical support document following DOE rating calculation methods Research findings indicate that actual installed energy use is much higher than estimated following DOE methods – thus the estimated savings is inappropriately elevated. 24

25. Best Available Data: Not Always Comparable - Part 2 25

26. Coordination and Timing Challenges Example Clothes washer recycling – adding water energy nexus benefits –Initial CPUC feedback in early 2015 directed the removal of water energy benefits as not approvable –In September the CPUC D.15-09-023 included provisions and direction including water energy benefits This is an example of a challenge in keeping workpaper reviews on track but also allowing for the dynamic policy environment. 26