HP-Flex: Next Generation Heat Pump Load Flexibility

Develop and test an open-source building energy management system, called HP-Flex, that controls heat pump (HP) settings in small/medium commercial (SMC) buildings to provide load flexibility (LF)—shape, shift, shed, and/or shimmy—while meeting

Lawrence Berkeley National Laboratory


Berkeley, CA

Recipient Location


Senate District


Assembly District



Amount Spent



Project Status

Project Update

The HP-Flex software platform was successfully tested in the FLEXLAB testing facility at LBNL. Site designs are complete and procurement has begun, installation is set to occur in spring-summer 2024. Additional work is being conducted to extend the model predictive control capabilities to variable refrigerant flow systems. Market analysis interviews have been conducted and results are being compiled.

The Issue

Small and medium-sized commercial buildings (SMC) have a great potential for optimal load flexibility (LF) because the vast majority of these buildings in California have a similar space conditioning configuration that use multiple packaged heating, ventilation, and air conditioning (HVAC) units with relatively simple controls. While a few approaches exist for optimal control of SMC HVAC for energy efficiency and simple demand response (DR), there is no control solution for this market segment that can reliably and at low cost deliver LF capability for optimal load shift, shed, shape, and shimmy, while maximizing occupant comfort.

Project Innovation

This project develops and tests an open-source building energy management system, called HP-Flex, that controls heat pump (HP) settings in small and medium-sized commercial (SMC) buildings to provide load flexibility (LF); shape, shift, shed, and/or shimmy; while meeting occupant needs and minimizing operating cost. The system includes new optimization software and equipment interfaces that together optimize HP operation, while being extensible to manage additional equipment such as refrigeration, water heaters, electrochemical and thermal storage. HP-Flex's standardized, modular design aims to make it easier to configure than existing systems, thus enabling a more cost-effective and reliable control for SMC applications.

Project Goals

Development of the control platform HP-Flex for model predictive control of heat pumps in buildings.
Demonstrate the HP-Flex platform in at least 5 small to medium commercial building sites.
Integrate new HP hardware into an open-source middleware platform (Volttron) and a proprietary IoT platform.
Develop educational curriculum for engineering and building trades based on the lessons learned during the project.

Project Benefits

This project hopes to remove barriers to the widespread adoption of this newly developed HP-Flex system by using experience from this project's field tests to develop educational curricula to train future engineers and technicians who will design, install, and maintain these systems. This project targets heat pumps used for HVAC applications. Industry partners who participate in the TAC will help disseminate findings and may adopt the resulting software (HP-Flex). The HP-Flex package supports the emerging business model of optimization as a service; which lowers the barriers to entry by providing energy management on a subscription basis. The findings from the algorithm development and from the long-term site demonstration will improve understanding of LF capabilities of HPs, along with impacts on grid and building end users, which will inform the development of a guidebook for future installations.

Lower Costs


HP-Flex offers greater benefits to building owners and the grid by allowing the integration of several types of building equipment (e.g., HP space conditioning, HP water heaters, thermal storage), unlocking the flexibility in heat pump loads, optimizing equipment operation, and lowering deployment costs in real buildings. Estimated reduced peak cost of 20% annually and a simple payback period of less than 2 years.

Greater Reliability


Estimated peak demand reduction of approximately 25% and GHG reduction of approximately 40% for the SMC sector, assuming a operating schedule from 7am to 7pm and reduction in nighttime load.

Key Project Members

Richard Brown headshot

Rich Brown

Research Scientist
Lawrence Berkeley National Laboratory
Project Member

Marco Pritoni

Research Scientist
Lawrence Berkeley National Laboratory


grade Tech Partner

Cal Poly Humboldt Sponsored Programs Foundation


Western Cooling Efficiency Center - UC Davis


County of Los Angeles


Los Angeles Community College District

grade Tech Partner

Quantum Energy Services & Technologies, Inc. (DBA: QuEST)

grade Tech Partner

MelRok, LLC


BlocPower LLC


Bakersfield College

Match Partners


Lawrence Berkeley National Laboratory

Contact the Team