For the September issue of Aamidor Consulting's Smart Building Insight newsletter, we interviewed John Powers, Founder and CEO at Extensible Energy. An except of the interview appears in our newsletter and the full discussion is here.
You can subscribe to our newsletter here, and learn more about Extensible Energy at their web site.
Q: Tell me about Extensible Energy, and your positioning in the market.
A: Extensible provides commercial buildings with an AI-powered software solution to manage HVAC, solar, and batteries in real-time while automatically decreasing energy costs and carbon emissions. In one sense, we’re like a BrainBox or EcoPilot, but we call ourselves an energy control platform. ‘Platform’ gets overused, but we’re going with it because from day one, we’ve designed our solution to go beyond HVAC. We have the ability to connect with and manage all the key energy loads in a building, including solar, EVs and batteries. And, our software enables buildings to be grid-interactive, generating new revenue from grid-services like demand response.
With over 10,000 days of operations across a very diverse portfolio of buildings we’re seeing dramatically reduced energy and demand charges, along with all the other benefits of predictive control like improved comfort, extended equipment life, and lower emissions. And, because we’re cloud-based our clients also get remote access and control from a single interface for their energy, climate, and emissions data over an entire portfolio of different equipment, and that also enables them to perform remote fault detection and diagnostics and reduce truck rolls and service calls. One customer called us an “energy focused BMS-lite… that actually takes action”.
Q: John, you've been in the industry for a long time - perhaps you can provide a bit of detail on your background?
A: My first job out of college was in the rates department of an electric utility. So, I learned how utilities make rates, and I know a lot about wholesale and retail utility electricity pricing. From there, I went to Quantum Consulting and helped grow that from three people up to 150 working on load research, demand side management, program evaluation for demand response programs and some software development. That software development led to a company called Energy Interactive, where I was the CEO and co-founder of a software business that helped utilities to engage with their customers. We sold that to ABB. I eventually left ABB in 2002 and started consulting again. Eventually, my consulting experience led me to realize that buildings could be a huge distributed energy resource if we could control the building’s flexible loads. And, we realized that building controls were pretty much stuck in the 1990’s, with a huge opportunity to take advantage of emerging technologies like machine learning and AI. That’s when Extensible Energy transitioned to a software company. We got a grant from the US Department of Energy in 2016, which allowed us to develop a working prototype of DemandEx, our core software, which we started commercializing in 2020. In 2021, we received seed-round funding from Blu Ox Ventures, which has been integral to helping us to bring DemandEx to the market.
Q: What’s Extensible’s big vision?
A: It’s audacious, but we see a very unique opportunity to tackle two very big problems with one solution.
We see the possibility of transforming every building into a dynamic energy resource, while also making it massively easier for building operators to manage their assets. And it’s a big opportunity. The energy transition is going to fundamentally change how we operate our grid and how we price our energy. Because renewable energy is intermittent, the need for flexible capacity is skyrocketing. One study estimated that the US will need the equivalent of 3,000GW of 2hr batteries to replace the capacity value that used to be provided by fossil fuel generators. We see every building as the equivalent of a 2hr battery. For example, if you can control just 50kW in 100,000 of California’s commercial buildings, that’s the equivalent of a 5GW, 2hr battery. That’s a pretty valuable asset that can be harnessed without having to build billions (trillions?) in new infrastructure.
At the same time, there are piles of studies and surveys pointing out how many buildings are poorly controlled, which means they are exceedingly hard (translation - expensive) to manage, there are uncomfortable people and lots of wasted energy. Everyone knows these stats, but they are still worth repeating: 30% of energy is wasted, 90% of buildings fail the ASHRAE comfort standard, 87% of buildings lack any building automation systems. Legacy systems are still incredibly expensive, and few, if any of them, are user-friendly and have real optimization. Their focus remains on delivering pre-programmed, static schedules. So we saw an opportunity to piggyback on modern technologies like cloud-computing, cheap wireless sensors, mobile apps, machine learning and AI. Put these together, and you can make a step-change improvement in building controls at a radically lower cost.
Q: What were the hard problems you had to solve that you didn’t already see out in the market?
A: Well, there were several intertwined problems. First, we had to demonstrate optimization algorithms could effectively manage comfort, while also substantially reducing peak loads and energy use. Then, tackling the challenge of demand charges was not trivial at all because it required a whole new level of real-time coordination and control. Demand charges are set by your highest 15-min energy use each month, and can represent up to 70% of a customer's bill. Our predictions were pretty good, but no prediction is perfect. So, we had to develop the ability to react to unexpected events in less than 2 minutes, and take immediate action to avoid setting a peak.
And our DOE research was forward looking to a world where most buildings have solar, EVs, batteries, AND those buildings were going to be grid-interactive, which means you have to optimize dynamically in real-time. That required us to take a very different, whole building approach that did not really exist yet. We couldn’t optimize each building equipment system in a silo, we had to manage all of them, in coordination, with our optimization focused on the utility meter. The meter is what determines the customer’s utility bill, and the meter is where the utility measures performance for payment of grid-services like demand response. So we had to develop new techniques for forecasting on-site solar production, and how to react to a cloud passing by. We learned how to manage the throttle on a battery, and we’ve demonstrated the ability to manage EV charging… all in coordination to deliver results at the utility meter.
Finally, we knew we had to build a platform that was both very low cost and easy to install and a back-end that was highly scalable. Our installed equipment is just a small gateway and energy meter. If a client has a BMS or smart thermostats we can often drop the gateway and connect cloud-to-cloud. That’s fast and cheap. We don’t talk about the back-end infrastructure much because it isn’t one of those ‘sexy’ topics, but without it, none of our fancy algorithms would be worth much. What it enables us to do is quickly add connections to new hardware in weeks, not quarters, and we can quickly test and deploy new features and algorithms without ever visiting a site.
Q: In many cases, different solutions are focused on different industry/building types, or different sizes of buildings - what is Extensible's focus?
A: We’re serving the non-residential building market, from 25,000 sq ft to 250,000 sq ft. That typically means schools, offices, retail stores, conditioned warehouses, and government buildings. We’ve been focused on the US market, but have our first building in Canada this month. Portfolio owners, in particular, see additional value from our software because of its ability to provide centralized and remote visibility and control over all their buildings and different equipment. We’ve yet to find a company that has the same equipment in every building. In reality, companies are always building, buying, selling, and upgrading all the time.
Q: You need a BMS or smart thermostat to deliver your solution - doesn’t that really limit you?
A: Well, yes and no. There’s an interesting report from James Dice and Nexus Labs that reveals how 87% of small to medium buildings have no smart controls, not even a smart thermostat. His research revealed that traditional BMS and EMS systems are too complex and expensive for these buildings. So, none of those buildings are optimized, and that’s a huge opportunity to unlock energy savings and also bring the benefits and convenience of a BMS to an unserved market. But, you have to be able to deliver that value at the right price point - about 1/10th the cost of a traditional BMS. We designed our platform to be cloud-based and work with very low cost IoT devices because we want to serve this massive untapped market. If a building has a BMS or smart thermostat, we can typically install in a day. If they don’t, the savings from our platform (or utility incentives) often cover the cost of upgrading their building with smart wireless controls, unlocking all the other benefits of having a building management system. Typically, these building owners are ecstatic for the convenience and benefit of modern technology.
Q: While many energy management solutions focus on saving energy, differing approaches mean that there could be more partnerships than currently exist - are there peers or complementary firms that are particularly good partners for Extensible?
A: Absolutely. We’ve found that HVAC control contractors are very interested to introduce their customers to us because the savings we generate becomes capital for system upgrades, and our real-time data enhances the level of service they can provide to their clients. We’ve also done a lot of work with solar installers. Solar is great for reducing energy charges, but it doesn’t reliably reduce demand charges caused by peak loads. Those peaks can be caused by a cloud passing over, and they often occur on the shoulder hours of the day when solar production is low. And, our software delivers the same peak shaving benefits as a battery, but at 1/10th the cost, and no concerns about fires. So solar installers bundle our service and the extra savings increases project ROI by as much as 50%. We also work with energy services companies (ESCOs). The energy and emissions reduction we can deliver are often as large as the deep energy efficiency retrofit projects, but we can get installed in a day without disruption to the facility.
Q: One key value proposition for your technology is that with more dynamic energy rates, and more renewables on the grid (and at customer sites) there are more opportunities to optimize energy use and save on demand charges. Perhaps you can provide some detail on your thinking here?
A: Yeah, you nailed it. Twenty years ago, almost your entire utility bill was driven by how much you used - $ per kWh. That’s changed dramatically in just the last five years, because utilities are using time-based price structures to grapple with the challenge of intermittent renewable energy. We see it in different forms, but it’s all time-based: demand charges, coincident peak charges, time-of-use charges. We’re seeing rate structures change everywhere from California and Texas, to Canada, the UK and Australia. We have seen customer bills where up to 70% of the total bill is determined by when they used power, not how much! That’s quite a paradigm shift from the way we’ve usually thought about managing energy costs.
Not only is the price of wholesale energy starting to get more volatile, the load profiles of our buildings are also changing and getting more unpredictable. If you add solar to your building, your load profile looks like ‘devil horns’, with peaks early in the day and in evening. If a cloud passes over your building, your meter can go from pushing a lot of energy to the grid, to pulling a lot of energy from the grid. Electrification of heating (heat pumps), batteries, EV chargers all dramatically change the shape and volatility of a building’s load profile. That’s where we thrive.
Q: What are some of the other tailwinds driving adoption in solutions like Extensible?
A: We’re seeing lots. Policy at the national level leans heavily in our direction, and this includes FERC 2222, which will take time, but it is driving the creation of a broad market for DERs to get compensated for grid services. Also the incentives of the Inflation Reduction Act and the electrification of heating and transportation. As buildings transition to heat pumps and install EV chargers, that just gives us more and more ‘energy under management’ and more value we can deliver to the building and the grid. And solar keeps getting cheaper, but savings behind the meter improve dramatically with automated controls that understand utility prices.
Q: Between the costs and potential savings, what's a typical ROI that a potential client could realize with your solution?
A: We like to think of this as a no-brainer for every building. Basically for a cost equivalent to a few hours of staff time, you get a massive upgrade to your building operations, lower energy costs, lower emissions, and your building becomes grid-interactive. Usually our clients see payback in less than one year, and then savings continue to accrue every month. We typically see 25% reductions in HVAC energy use, and up to 40% reductions in demand charges. Those are pretty easy to quantify. But customers also realize other savings from reduced operating and maintenance costs. One building engineer who manages 2m square feet across a portfolio said that Externsible would reduce his need for truck rolls and just the service call savings was a “no brainer”. Extended equipment life from smarter schedules also has a real value. We let the customer assign numbers to those benefits, but they are usually substantial, especially across a portfolio. We also help customers identify the optimal utility rate schedule for their building, and that can be another 10-20% savings. Sometimes, using a strategy we call ‘hard cap’, we can eliminate demand charges completely, by keeping a building from exceeding a kW maximum. Finally, we give customers all the data they need to meet the sustainability reporting requirements, which avoids the need to purchase other systems.