Space Conditioning Interactions with Heat Pump Water Heaters: A Test in the PNNL Side-by-Side Lab Homes

Space Conditioning Interactions with Heat Pump Water Heaters: A Test in the PNNL Side-by-Side Lab Homes Cheryn Metzger, Sarah Widder, Joe Petersen, Josh McIntosh Pacific Northwest National Laboratory February 27, 2017 Portland, OR PNNL-SA-124136

BACKGROUND

Definition: Interaction Factor (HC ƒ ) = Thermal Coupling as percent of the total theoretical HPWH thermal load that impacts space conditioning load In English: How much energy does the space conditioning system have to make up when a HPWH is in a home?

Motivation November 2014, the Regional Technical Forum (RTF) accepted a preliminary interaction factor (HC f )= 65% based on preliminary data from the PNNL Lab Homes. 1 The RTF also directed staff to develop a research plan to study the HVAC interaction effect induced by heat pump water heaters (HPWHs), which was accepted in April 2015. Research plan consists of two primary components: 1. Theoretical analysis (Ecotope) to describe basis for HC f 2. Experimental study in the PNNL Lab Homes to determine/validate range of expected field HC f values 1 Widder, et al. 2015. Impact of Ducting on Heat Pump Water Heater Space Conditioning Energy Use and Comfort. PNNL-23526. Pacific Northwest National Lab. Richland, WA.

Theoretical Interaction 1. Some of the heat from the space is latent heat (Q HPWH,latent ), which will not affect the ambient dry bulb temperature of the space 2. Localized cooling may decrease the heat transfer from the space as compared to the pre-retrofit case (UAΔT HPWH < UAΔT no HPWH ) 3. Some of the heat imparted to the water heater from the space may be free heat from solar gains that would also decrease the total heat transfer from the space if the house drifts above set point 4. HPWH near a thermostat may artificially call for more heat than necessary for thermostat to maintain set point

Temperature ( C) HPWH Air Inlet Temp (F) Past Results Localized Cooling 70 Lab Homes show localized cooling in water heater closet of ~8.4 F (~4.7 C) 65 60 55 50 45 40 (12/18/13 00:00:00) (12/19/13 00:00:00) (12/20/13 00:00:00) (12/21/13 00:00:00) Timestamp LHA - Fully Ducted LHB - Unducted NRCan shows localized cooling in basement of ~3.6 F (~2 C) 21 20 Experimental House Basement Temperatures Basement S Low Basement W Low Basement NW Low Basement N Low Basement NE Low 19 18 17 16 15 30-Dec-11 0:00 29-Dec-11 12:00 29-Dec-11 0:00 28-Dec-11 12:00 28-Dec-11 0:00 27-Dec-11 12:00 27-Dec-11 0:00 26-Dec-11 12:00 26-Dec-11 0:00 25-Dec-11 12:00 25-Dec-11 0:00 24-Dec-11 12:00 24-Dec-11 0:00 23-Dec-11 12:00 23-Dec-11 0:00 Date and Time

TEST PLAN

Research Strategy: PNNL Lab Homes Research Goal Observe the space conditioning energy interaction that results from the installation of a HPWH in interior spaces Data Collection PNNL Lab Home Study: Test four install locations throughout home Analysis Use a HPWH on a dolly operating on a known high load profile compared to a ERWH operating in the baseline home Compute change in space heating requirement, change in latent heat across the coil, and any impact on localized cooling to compute HVAC interaction factor for each location Observe range of results

Experimental Plan A B C D Test Case Description Test Location Reason to Include Test Case Master Bath, Door Closed A Most disconnected to thermostat, but still in conditioned space Utility Closet B Most connected to the return duct Living Room C Most connected to the thermostat Water Heater Closet D Represents semi-conditioned space 9

EXPERIMENTAL SETUP

Side-By-Side Homes Are Fully Equipped Energy metering 42 individually monitored breakers with ½ controllable Temperature and relative humidity 15 interior room temperature thermocouples Water and environment Controllable water flows at fixtures On-site weather station Data collection via 2 Campbell Scientific data loggers/home 1 minute data, sometimes averaged over longer periods Per Home!

Monitored Data Around Water Heater *Note: Picture demonstrates monitored values only. Location of sensors is not necessarily accurate. 3/6/2017 12

Fully Mobile Water Heaters Air Flow Sensor Bracket Tipping Bucket Flexible Hoses for Water Inlet and Outlet Solar shielded ambient T/RH sensor Rolling Platform Plywood False Floor March 6, 2017 Josh McIntosh, Joe Petersen, Greg Sullivan 13

Close Up of Measurements Two T/RH Sensors (Campbell Scientific HC2S3, 75ft cable, ±0.1 0.8% RH). Top arrow points to supply T/RH measurement, bottom arrow points to exhaust T/RH measurement. Condensate Drain Tipping Bucket (Rainwise Inc., Rainew 111 Single Counter. 0.01 inches per tip. 2% accuracy at 1.5 inches per hour) 3/6/2017 14

Utility Room Setup 3/6/2017 15

INITIAL RESULTS

General Observations Previously mentioned studies could be considered semi-conditioned space (garage, basement, water heater closet). In current experiment, water heaters are truly interacting with the thermostat in conditioned space. Water draw profile is high (84 gallons per day) compared to national or regional averages in order to capture the high end of a realistic signal. Heat pump only mode so no input energy coming from resistance elements. All input energy from either the air or the compressor motor. Thermostat set at 71 F, water heater set at 125 F 3/6/2017 17

Whole House Energy Consumption (Watts) Whole House Energy Use During Calibration (1/15/17) 12,000 10,000 8,000 6,000 4,000 2,000 0 12:00 AM 4:00 AM 8:00 AM 12:00 PM 4:00 PM 8:00 PM 12:00 AM Time Baseline Home Experimental Home March 6, 2017 18

Experimental Plan A B C D Test Case Description Test Location Reason to Include Test Case Master Bath, Door Closed A Most disconnected to thermostat, but still in conditioned space Utility Closet B Most connected to the return duct Living Room C Most connected to the thermostat Water Heater Closet D Represents semi-conditioned space 19

Energy (Wh) Outdoor Temperature ( F) Difference in Daily Water Heater Energy: (Baseline minus Experimental Home) 14000 40 12000 35 10000 30 8000 6000 25 20 15 4000 10 2000 5 0 Baseline (ER) WH in Mstr Ba WH in Liv Rm WH in Utl Rm 0 Δ Daily Water Heater Energy (Wh) Average Outdoor Temp ( F) 3/6/2017 ER = Electric Resistance Mode 20

Energy (Wh) Outdoor Temperature ( F) Difference in Daily HVAC Energy: (Experimental minus Baseline Home) 35000 40 30000 35 25000 30 20000 15000 25 20 15 10000 10 5000 5 0 Baseline (ER) WH in Mstr Ba WH in Liv Rm WH in Utl Rm 0 Δ Daily Heating Loads (Wh) Average Outdoor Temp ( F) 3/6/2017 ER = Electric Resistance Mode 21

Temperature ( F) Experimental Home - Hall Temperatures in the Experimental Home 75 70 65 60 55 50 Thermostat doing a good job at keeping the hall temperatures fairly constant so we can compare results 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hours of the Day Baseline Experiment (Jan 15) Master Bath Experiment (Jan 20) Living Room (Jan 26) Utility Closet Experiment (Feb 1)

Temperature ( F) Temperature ( F) Local Cooling Effect in Small Spaces Local cooling is about 17-20 F in small spaces 75 70 65 60 55 50 75 Master Bathroom Temperatures 1 2 3 4 5 6 7 8 9 101112131415161718192021222324 Hours of the Day Utility Room Temperatures Calib Exp Home Calib Base Home Mst Ba Exp Home Mst Ba Base Home 70 65 60 55 50 1 2 3 4 5 6 7 8 9 101112131415161718192021222324 Hours of the Day Calib Exp Home Calib Base Home Utl Rm Exp Home Utl Rm Base Home 3/6/2017 23

Temperature ( F) Master Bedroom Temperatures Increased space temperatures in master bedroom - potentially due to local cooling around thermostat Solar gains because master bedroom on south side 75 70 65 60 55 50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hours of the Day ERWH Calib Exp Home ERWH Calib Base Home HPWH in LR ERWH in LR HPWH in Util Rm ERWH in Utl Rm

General Observations Theoretical interaction factors assume a single, well-mixed zone miss most of the relevant processes. Energy use in home with HPWH increases as HPWH is closer to the thermostat and/or return grille Local cooling effect is more than double of previous experiments in small spaces. Conditioned space, not ducted to the outside. As expected, indoor winter conditions are dry so no condensate measured HVAC load due to HPWH interaction is exaggerated in this instance March 6, 2017 25

Future Work Flip flop homes. Put water heater in HPWH mode in baseline home and in ER mode in experimental home. Compare results to previous experiments. Additional experiment still needs to take place in the water heater closet. Results can be compared with previous experiments in the water heater closet. Determine if stratification is occurring in each room, and contributing to elevated temperatures RTF contract analysts have requested a space heating experiment using similar approach, but with space heaters instead of HPWHs. This may shed light on sensible heating interaction vs. sensible cooling interaction Report all results in HCf March 6, 2017 26