Load Flexibility and Rate Design Considerations for Heat Pump Water Heaters William Burke Deepak Aswani Sacramento Municipal Utility District www.peakload.org
About Sacramento Municipal Utility District • Not-for-Profit Municipal Electric Utility • Sacramento County in California with about 625,000 accounts • Vertically integrated with generation, T&D, and customer facing functions • Net zero carbon goal by 2040 impacting the generation portfolio, appliance electrification, and load flexibility initiatives www.peakload.org
Present day opportunities for Time-of-Day Rates • SMUD Residential Time-of-Day Rates helps align customer costs with utility costs on average • Provides an opportunity for shared savings Actual Utility energy • However, cost of energy and capacity Day 1 peak Day 1 off-peak or capacity costs vary day to day Day 2 peak Day 2 off-peak (bulk and local) are more complex Day 3 peak Day 3 off-peak than Time-of-Day captures … Day n peak Day n off-peak www.peakload.org
Load flexibility with Heat Pump Water Heaters • Changing the water temperature setpoint to overheat the water allows for load shifting • Mixing valve keeps temperature constant for consumed water • Shift load without the customer noticing! www.peakload.org
Looking beyond Time-of-Day Rates Transactive Flat monthly Dynamic Time of Day energy (peer bill (cell rates rates to peer phone plan) trading) Greater Greater utility customer responsibility responsibility Traditional Volumetric rate with rates load shift credit Aspiration for a simplified customer billing contract that also provides utility load flexibility www.peakload.org
Pilot design • Research compares 3 modes of HPWH operation 1. Baseline – no active load shaping 2. Time of Day (TOD) rate optimization – load shape to reduce on-peak consumption and increase off-peak consumption 3. Utility influenced load shaping based on utility day ahead electricity costs • Customer contract design • Default mode is TOD rate optimization • 120 days/year can be event days – either Baseline or Utility load shaping • Customer receives $75 up front credit for signing up for the program and $2/month participation bill credit www.peakload.org
Founded: 2014 Located: Louisville, KY Business Model: SaaS Other Programs: ✓ Glasgow/TVA SET Program (Battery, HPWH, Thermostat) ✓ GMP BYOD Program (Battery) ✓ Belmont Light Peak Reduction Rewards (All Devices) ✓ GMP EV Unlimited Program (EVSE) www.peakload.org
Broad Manufacturer Support Water Water Room AC & Room AC & Thermostats Thermostats Storage Storage EVSE EVSE Heaters Heaters Mini Mini -Splits Splits Robustly integrate with a new device in ~2 weeks www.peakload.org
with • Water Heater Optimization Program BYOD Enrollment • Multi Manufacturer Support • www.peakload.org
Rapid Algorithm Development Software in the Loop Simulation Explanation • Models Device Behavior • Models Human Behavior • Operate Actual Control Software on Simulated Devices Benefits • Accelerate Time • Test With No User Disruption • Replicable experiments • Faster Development • Lower Risk www.peakload.org
TOD Rate Optimization • Learn Water Use Behavior Read from Price Server • Take Advantage of Repeating • Schedules Adjust Schedule to Save • Money Ensure Water Delivery • Temperature www.peakload.org
Wholesale Cost Optimization Water Heater Arbitrage • Learn Water Use Behavior Integrate with Pricing Server • Dispatch Based on Hot Water • Capacity and Price Optimize Wholesale Cost • Ensure Water Delivery • Temperature www.peakload.org
Impact of TOD optimization • 22 customers (9 GE Geospring 13% reduction and 13 Rheem) vs. baseline • TOD optimization implemented so far • Utility influenced load shaping going live in Fall 2019 www.peakload.org
Program Challenges • Mixing Valve Support Safety Concerns • • Terms and Conditions Mixing Valve Safety • • Multi-Manufacture Support Different min and max temperatures • Different data granularity • • Population Size Learning benefits from large population • www.peakload.org
Customer Learnings • Simple Enrollment a must • Few People Change Behavior • Fixed Schedules Don’t Work Unhappy Customers in • Control Group Learning Solved the Problem • www.peakload.org
Recommend
More recommend
