Progress 09/01/23 to 08/31/24

Outputs
Target Audience:Our target market is small to mid-sized commercial buildings. Our customer decision-makers are superintendents of business, COOs, building owners, facility managers, and maintenance directors for building portfolios, typically with 50,000+ square feet under management. > BEACHHEAD MARKET SEGMENT: K12 SCHOOLS CEL has market knowledge, customer insights, and traction in the K12 segment. There are about 100,000 K12 schools in the US. They are the largest energy consumer in the public sector, spending more on energy than on computers and textbooks combined. HVAC energy accounts for 46% of their total energy consumption. Although there is a great potential for energy savings, cutting costs in this area risks creating an uncomfortable learning environment for students and other occupants. CEL's autonomous technology maintains occupant comfort and optimizes building operations to respond to utility grid signals, demand charges, and time-of-use prices that significantly impact electrical bills. Without CEL's technology, users' already-high energy costs would increase. > MARKET SIZE Total Available Market: Buildings in the Government, University, and Schools (GUS) subsegment of the market make up 28% of US commercial floor space. With lower-cost options like CEL, the market opportunity for energy savings investments in the GUS market is $66-208 billion, and the potential energy savings of all GUS building projects is 200-262 trillion Btu. Served Available Market: Education accounts for 7% of total commercial buildings and 14% of commercial floor space in America, but ranks third in total floor space and second in average floor space (31,200 sq ft). Meanwhile, only 10% of schools are to have building controls - leaving the market open. CEL estimates a total market opportunity in K12 of $9-$30 billion. Serviceable Obtainable Market: Given CEL's large market opportunity, Obtainable Market for the next several years will be constrained primarily by our ability to scale due to factors like the time required to create market awareness, the practical limits of adding and training personnel, available growth capital, our ability to maintain a competitive advantage, and expanding our geographic footprint from the US to a worldwide market. It is not unrealistic to expect long-term market penetration of 10% to 15%, representing a greater than $1 billion opportunity. > CEL'S ADVANTAGE: CUSTOMER DECISION TO PURCHASE When evaluating CEL's solution, building operators will be sensitive to up-front cost, payback complexity, sales process length, disruption to operations, and ease of use. Imagine a hypothetical Superintendent for a K12 school district in California. Her utility just changed their prices--resulting in a 40% energy bill increase--and the school board mandated all facilities be carbon-neutral by 2025. What are her options? She might be spending $3M per year on energy, but a building energy management system can cost hundreds of thousands up front, require she pass a bond measure or navigate procurement, and disrupt the use of the school during the extensive install. Her maintenance crew is ill-equipped to deal with the complexity of traditional solutions. So, like most middle- market customers, she skips the automation system altogether. Instead, she manually tracks energy costs in a spreadsheet and sends crews to program hundreds of devices by hand. This is a typical scenario that CEL found in its customer discovery process. In 2020-2021, CEL conducted multiple levels of in-depth customer research and validation. Through CleanTech Open and WSU's NSF i-corps program (award #1547873), CEL interviewed over 100 potential customers. In 2021, through a project with the Electric Power Research Institute, Southern California Edison, and Silicon Valley Clean Energy, CEL conducted an average of six hours per person of in-depth user experience research. This was done with over 30 facility managers, architects, maintenance directors, and energy managers from municipalities and school districts. Schools said they wanted something that was affordable, simple, and hands off that works. CEL discovered that in the public building sector EMS end users tend to be maintenance staff or service contractors. Average procurement thresholds for public buyers under which a single person can authorize a sole source procurement vary significantly; however, the majority of buyers we interviewed are able to make $25k-50k purchases without higher-level approval. Our price point ensures that a CEL salesperson can make a direct sale of a demonstration site through a sole source procurement contract in under two months and then justify "value for money" expansion through demonstrated energy savings at the initial pilot site. Changes/Problems:Risk factors for this project include slow procurement processes, complex stakeholder hierarchies (even for smaller ticket items), and implementation schedules that are driven and constrained by building occupancy schedules. Procurement: CEL has identified a staged customer onboarding and success process where we present board memos, templates for decision makers and FAQs that will allow them to quickly gather the necessary diligence to execute a sole source procurement. As a precursor to this project our team has defined each step of the onboarding process and defined roles and responsibilities for ensuring each new customer receives the requisite paperwork and supporting documents as soon as possible. Complex stakeholders: CEL is increasingly attempting to identify all relevant stakeholders withpotential to stop or slow down projects at the onset and to address objections prior to establishing an installation date or procuring equipment. We are developing FAQs and best practices for a wide variety of stakeholders from boards to finance directors to HVAC technicians. We increasingly have a list of staged installation documents that engage key stakeholders such as IT at the earliest possible moment to ensure that they have buy-in and are ready once installation is scheduled. Implementation schedules: K12 in particular prefers to schedule installations during school breaks, which also roughly coincide with times when families and the trades want to be on vacation. This has created problems in the past when we are unable, in some areas of the country, to find licensed electricians to install meters during times when the schools want to have their systems put in place. CEL is building a network of qualified contractors throughout the West Coast on whom we can call in future installation. Our customer success team is building a series of training resources that can help us source, train and reduce onboarding time for licensed contractors in regions where we heretofore have not completed an installation. We have now also reduced our average installation time to less than 8 hours on even complex installations using a new installer. We expect to continue to increase our install efficiency as we document and repeat processes used in the past. We will also look toward means of batching installs geographically to maximize efficiency during times when many schools might be on break at once. Mechanical & electrical systems well beyond end of life We have found that rural sites require significant building equipment upgrades. Most equipment at the locations we encountered is over 20 years old. Some are part of brittle and aging hydronic systems that are in a state of constant emergency (springing leaks that cause water damage to buildings and grounds). Maintenance of these systems costs nearly 4 times that of sites with simpler HVAC configurations, leaving little money, resources, and time for upgrades. Even for those rural communities that wish to update their HVAC systems to be more efficient, all-electric systems often are hampered due to limited electrical capacity in crowded and aging electrical infrastructure. In addition, the decline of many rural economies (such as fishing and timber) has led to a declining revenue base in their communities, limiting their ability to finance overhauls using bonds. During a site visit to a city hall that is located in rural Oregon and was built in the late 1800s, we discovered that the building had multiple generations of HVAC and electrical equipment installed. Decades of office redesigns and iterations created conflicting zones that were served by differing air handling units and heat pumps throughout the building. This disjointed and complicated system layout added to the complexity of creating a proper metering strategy for the building in this rural location. The communities where we installed equipment in Oregon, Washington and California are experiencing changing and extreme temperature fluctuations that occur in both the summer and winter seasons which exacerbate the maintenance burden for their lean staff. For example, CEL postponed a potential site visit when coordinating with Wenatchee school district, a community in Washington, due to a severe cold snap and ice storm that ruptured water pipes in multiple buildings. Maintenance and operations teams are under constant pressure to keep building infrastructure running through these extreme weather events. Controls systems are non-existent or outdated As schools electrify, it means they typically will add HVAC units that use electricity to heat and cool rooms in buildings. In the winter many of these units will begin running early in the morning to ensure rooms are warm when occupants arrive. In the summer and the fall the units will run in the afternoon to combat the impact of warm temperatures as the sun rises. Because many units are running at once it can cause an electrical peak that drives up cost and therefore necessitates smarter controls that reduce and shift peak demand. In most cases controls are installed when the HVAC system is fully or partially replaced. As mentioned previously, many of these systems in rural communities are close to two decades old. Thus we found many HVAC systems where controls either do not exist, or are running on PCs with a Windows 98 operating system (which is no longer supported by Microsoft). All other problems and corresponding solutions contain confidential information and have been fully disclosed in the USDA Interim Technical Report. What opportunities for training and professional development has the project provided?NYSERDA Entrepreneur-In-Residence (EIR) Program - 5 skills labs that CEL employees participated in that include: Customer Discovery, Sales, Investor Readiness, HR and Compliance, and Grant Writing. Nevada Climate Justice Convening - Collaboration between Desert Research Institute, Dream.org, and the US EPA to make meaningful progress through climate justice initiatives. CEL Product & Engineering Off-Site Workshop - Collaboration between CEL Product, Engineering, and Hiring Teams to steer product and engineering from strategic implementation to tactical. We prioritized updating our backlog, estimating resources, and improving communication from product & engineering teams to various audiences: investors, customers, future employees, and current non-technical teams such as recruiters and customer success. How have the results been disseminated to communities of interest?To ensure the alignment of this project with the needs and vision of the community, CEL adopted a hands-on approach to solicit community input with every piloted or designed project. Part of this process is organizing human-centered design workshops (for the design phase) and Smart & Resilient Schools (SRS) Summits (for school projects). We define a community as the collection of decision makers, stakeholders, end users, operators and occupants surrounding our customers. This project will include such workshops for each district included. > HUMAN-CENTERED DESIGN WORKSHOPS These interactive workshops engage diverse community members with unique perspectives in providing project feedback. One example is an EPIC project at Jamboree's Heroes Landing apartments for low-income, disabled veterans in Santa Ana, CA. Eight key community members (Asset Managers, Maintenance Directors and Technicians, Property, and Project Managers from Jamboree) participated in the three-hour workshop. A second workshop engaged a dozen residents in conversation about community energy usage patterns within their buildings and their needs when interfacing with emerging energy prices and technologies. > SMART & RESILIENT SCHOOL SUMMITS (SRS) In 2023, two rural K12 districts hit hard by the opioid crisis partnered with CEL to host an SRS Summit in Gray's Harbor, WA, sparking widespread enthusiasm for new avenues to reach youth. Governor Jay Inslee and state senators attended via video conference to connect with educators and stress the importance of exposing students to energy jobs, STEM careers and learning opportunities using technologies like CEL's. Educators ranked the summit in the 90th percentile on continuing education evaluations. Attendees included educators, administrators, facility professionals, and experts in energy, construction, and financing. Insights gained in workshops and SRS shape the design of CEL's platform to better meet the needs and expectations of the community. Post-event, design teams were inspired to develop flyers, training and educational content. This community engagement approach fosters constructive dialogue that is instrumental in aligning our projects with broader community needs and vision and incorporating community feedback in the project. CEL's project partners include Montebello Unified School District, Antelope Valley Union High School District, Pomona Unified School District, Temecula Valley Unified School District and Fremont Union High School District. All of these districts have waitlisted schools to try our tech. Three are in CalEnviroScreen-designated Disadvantaged Communities (DACs). Priorities for these communities focus on reducing energy costs and maintaining indoor air comfort for teachers and students. What do you plan to do during the next reporting period to accomplish the goals?We will continue to address the following customer needs: Reduce energy costs: American K12 school districts spend over $6 billion annually on energy. Up to 30% of a district's total energy goes to waste. Find an affordable solution: A typical building energy management solution would cost an average US primary school $125k-$250k, which is out of reach for most rural schools. Optimize for time-varying energy pricing: Half of US investor-owned utilities have introduced time-varying prices that encourage customers to use energy when it is cheap, plentiful, and emissions-free. This requires public, low-income buildings to adopt technologies with advanced controls, or risk facing a 40% higher electrical bill. We will continue our work on Objective 1: Test Prototype in Full Field Deployment, Objective 2: Verify Performance, and Objective 3: Commercialization for the next reporting period. Objective 1: Test Prototype in Full Field Deployment Task 1.2 Carry out Field Demonstrations Task 1.3 Commission, Train ML, Debug Objective 2 - Verify Performance Task 2.1 Develop & implement testing framework Task 2.2 Verify system performance Milestone 2 - Initial commercialization assessment Objective 3 - Commercialization Task 3.1 Develop effective onboarding & retention Task 3.2 Value engineer for cost efficiency Task 3.3 Integrate with pricing and marketing Milestone 3 - Sales material and targets

Impacts
What was accomplished under these goals? The SBIR Phase II project, led by Community Energy Labs (CEL), focuses on developing and commercializing affordable grid-interactive efficient building (GEB) HVAC control technology. The technology retrofits public, low-income Small and Mid-size Commercial Buildings (SMSCBs) and has completed field demonstrations at 14 diverse sites across urban, suburban, and rural locations. The project has shown promising results: it has reduced demand peaks from electrification by 11-30% and overall heating/cooling power by nearly 30%, all while maintaining occupant comfort. The technology's impact has been particularly significant in rural areas, where energy efficiency is critical due to higher electricity and heating costs. In-person training for building occupants has enhanced technology adoption and effectiveness while mitigating the challenges that rural regions face in finding skilled trade workers. With two rounds of Measurement & Verification (M&V) testing completed across seven sites in 2023/2024, CEL continues to refine the technology's scalability and has developed repeatable onboarding and customer retention processes to further accelerate adoption. This project addresses critical needs for energy efficiency in aging SMSCBs, supporting electrification goals and providing solutions to mitigate climate change impacts. CEL installs wireless sensors and equipment controllers and uses cloud-based software powered by machine learning algorithms to autonomously predict and efficiently control how and when building HVAC equipment is operating. CEL offers building owners a user-friendly interface with a back end application that reduces energy bills, promotes beneficial electrification, and improves resilience for existing building owners. We identified 14 sites from CEL's customer base that allowed us to perform field test demonstrations: in 3 Climate Zones (4C (Marine), 3B (Hot-Dry), and 3C (Marine)), with 3 Utility Tariff Structures (Fixed Rate with Demand Charge, Time of Use, and Dynamic rates), on 6 HVAC System Types (Rooftop Air Conditioner and Gas Furnace, VAV with Hot Water Reheat (Chiller/Boiler), Packaged Terminal Heat Pump, Ductless Mini-Splits Heat Pump, Rooftop Air Source Heat Pump, and Ducted Heat Pumps), and with 2 HVAC System Topologies (Single (1:1) and Complex (1:N) zoning). We installed CEL's Gateway at all 14 identified sites for field test demonstrations.We installed a Current Transformer (CT) on different electrical panels and subpanels to collect power data at the site and equipment levels. These are off-the-shelf commercial equipment that we use such as Egauge and Elkor meters. We also upgraded the thermostats at 2 of the 14 sites to support networked communication. Pelican is a networked thermostat manufacturer. We worked with the manufacturer and licensed HVAR/R technicians used to upgrade thermostats at Triangle Lake Charter School in rural Oregon and Central Elementary in the Hoquiam School District in Washington from non-communicating thermostat to fully networked models. The technology allows us to interact with schools' HVAC systems remotely and to perform M&V tests via Application Programming Interface (API) calls from our system. Measurement and Verification (M&V) Plan: CEL, along with UC Berkeley, designed an M&V protocol in order to gain a scientifically valid assessment of model performance in the field. The M&V protocol compares the baseline, or standard sequence of controls, against the intervention, CEL's hybrid MPC+ML controls. Instead of a typical M&V approach that compares "pre-installation" to "post-installation," the A/B testing framework allows us to create a truly randomized sample. For this reason, we chose to follow a randomized block scheme. This enables the M&V protocol to randomly select for each day whether to implement the baseline or the intervention, which is an efficient way to avoid bias and confounding factors. The randomization process will use 'blocking' to ensure that the same number of days are assigned to each control for each block period and will be repeated for several months. This approach can more accurately, fairly, and directly compare the performance between the two controls than the typical M&V approach because it allows for both controls to be tested across weather events in the same season. We were able to recruit/evaluate 14 sites for our field demonstrations, where we successfully completed site visits and installation of the monitoring, control and networking devices. Then, we proceeded to commission the sites. This step established read/write access to HVAC equipment controllers through our software services, where CEL developers were working with the Control System vendors for each customer to develop the appropriate equipment drivers and integrate them into our backend. During this process, we learned that some control vendors restrict third party write access to their systems. In order to gain that access, CEL engaged in establishing relationships with these vendors and working with their engineers on the necessary solutions that will allow CEL's services to have full write access to their control systems. Due to the nature of our technology that depends on seasonal temperature variations (cooling vs heating seasons), we proceeded with 9 out of the original 14 qualified sites, allowing us to accurately assess the impact of our technology on demand reduction and load shifting in both heating and cooling seasons. We conducted two rounds of Measurement and Verification (M&V) of the field deployment on the 9 sites where we established full read/write access to HVAC controllers between August 2023 and March 2024. After our 2 M&V trials of the 9 sites, the cleaned data that was analyzed represented 5 sites for the summer cooling season and 5 sites for the winter heating season, ending with a total of 7 distinct sites. The M&V rounds allowed us to study the performance of the control algorithms in reducing energy cost charges while maintaining thermal comfort as well as to test the reliability and scalability of our deployment prototype. The results shown below includes analysis from 2 different tariff structures: Demand Charges: charges based on how much power is used during a customers' peak power usage during a billing cycle. Dynamic Charges: charges based on when power is used and how the price fluctuates based on the time of use throughout the day. For demand charges, we measured and analyzed the peak demand reduction between the baseline and intervention controls, while also measuring the change in room temperature to ensure we are maintaining the customer's comfort. For dynamic charges, we compared how much energy was consumed at times throughout the day when energy is at a lower price to when it is at a higher price, again while measuring the change in room temperature. Peak demand reduction for customers with Demand Charges: Summer Season: 11-28.35% all while deviating from the room temperature constraints by less than 5%. Winter Season: 5-13% all while deviating from the constraints by less than 5% In the Summer Season, a load shift for customers with Dynamic Charges was shown to be as high as 14% from being consumed at a High Price Period to a Low Price Period.?

Publications