All Episodes

[Episode #16] – Energy Efficiency Markets


Improving efficiency is almost always easier and cheaper than generating new power, so efficiency should be our first target in energy transition. But it’s usually the last. And while there are very effective incentives for renewable energy, the incentives and programs for efficiency have been far less effective. In this episode we talk with efficiency guru and innovator Matt Golden about how to get away from efficiency incentive programs, and switch to performance-based markets for energy efficiency, plus how to standardize efficiency projects so that they are easier to understand, trust, and finance. Thanks to ideas like these, energy efficiency may be about to hit the big time.

Guest: Matt Golden, CEO of Open Energy Efficiency, and Director of the Investor Confidence Project

On Twitter: @GoldenMatt

On the Web:

Recording date: April 4, 2016

Air date: May 4, 2016

Geek rating: 9

Chris Nelder: Welcome Matt to the Energy Transition Show.

Matt Golden: Hi Chris.

Chris Nelder: So you've had a really interesting career path. When we first met around 2004 I was designing and selling solar PV systems in the Bay Area and you had a company that went around testing thermal and air leaks in buildings and helping people figure out how to correct them. In the 12 years since then why don't you tell our listeners a little bit about how your career has evolved and what brought you to where you are today?

Matt Golden: Yeah, well I actually started in the same place as you, one step even before getting into the efficiency business. I took a flying leap out of the technology industry and landed in solar back in the wild west, a very similar timeframe to when you were playing that game, and saw a gaping hole in energy efficiency. Really kind of believed the notion that there really should be an energy loading order and that there's this huge opportunity that really wasn't being tapped in energy efficient and the built environment. And as you said, I at that point made a jump out of the solar industry which was you know very much early stages, I was actually at a company called Sunpower and Geothermal Energy known as SPG, and we did the first megawatt in the state of California back when a megawatt was absolutely unfathomably big. We couldn't possibly believe how many panels it was right, at something like $9 a watt. So I watched that market evolve and actually taking a lot of cues from watching the solar industry mature from a bunch of, no offense, those hippies in NorCal selling solar on home equity lines, to the industry that its turning into today, which is really a different story and frankly energy efficiency, I hope we can learn some of the lessons from solar, but it is really destined to follow a somewhat similar path if we play our cards right. So interesting, you know it's kind of the Chinese proverbial definition of interesting times. But as you said I kind of jumped out of solar and kind of technology and got into what was really just a get down and dirty crawl space belly crawling contracting business doing home performance which was really energy auditing building science based whole house retrofitting in Northern California, and really grew that business from my third bedroom, the origin story I think with I sold my Westphalia to seed farm the company and we grew about 100 percent a year for about four or five years, just organically.

Chris Nelder: And that ocompany was Sustainable Spaces.

Matt Golden: That was Sustainable Spaces, which became later Recurve which rebranded after we took a few slugs of venture capital money and started becoming really two things. One, we started investing heavily in building software technology to try to improve a number of aspects of the business that actually continue to cause problems. And I'll talk about how we're actually I think just now coming up on solutions to some of those problems, but so we built tools for contracting companies to be able to go into a house, use a tablet PC, was not a good omen but we missed the iPad by about a year and half timing wise.

Chris Nelder: That would have been handy back then.

Matt Golden: It's all about timing isn't it. So we built a really powerful suite of tools, the technology is still in use that allows contractors to go into a house either tethered to the Internet or untethered, not only simulate the building and estimate energy savings across any number of energy conservation methods that we might employing, lightbulbs, HVAC, new duct work, et cetera, but also estimate the project which is maybe the hardest part of doing a business honestly is training a bunch of energy auditors to be able to be able to estimate seven or eight or nine different trades and actually deliver onsite even same day professional sales quality report to the customers so they can close the job and actually get itself built. Which is a minor detail when it comes to energy efficiency that we'll talk about more but there's the great promise of energy efficiency and then there's the barriers that stand in its way. And that's kind of what we're trying to overcome here. So Recurve which is what Sustainable Spaces became after again raising capital. The joke, and this is actually not a market timing joke, but we raised money on Friday and on Monday morning Lehman Brothers went bankrupt. So that was our market timing, talk about into a headwind. It was definitely a challenging run. So we were trying to scale into basically that collapse. An interesting set of circumstances. And really betting very heavily that the Recovery Act and these public programs are sprouting up all over the country because we had actually grown our business before there was a program in California. We were the first real home performance BPI accredited company and nobody was there to help. So we were betting on the industry taking a turn up and to the right based on all the money that was going into home energy efficiency, over $500 million dollars that was deployed to the Department of Energy into Better Buildings Neighborhood programs, there were utility programs spooling up all over the country. And what we found unfortunately was that we were writing the best software in the country and I think honestly we actually just won a software shootout seven years later... Somebody still has Recurve software running seven years later at an event, but we found we couldn't sell our software to contractors even though they desperately needed it. Did you buy it and actually bought it. Also having to buy whatever piece of regulated software the program required them to use and that we bumped into around the country, in California but really all around the country. What we hoped to become markets were really little regulatory fiefdoms. And so contractors had to use whatever software toll the program told them to use. That was a big challenge to the software business. And then you know the contracting business again it was just into a headwind, so the market conditions were challenging but when the programs arrived what we actually found is that they trained a tremendous amount of competition. We were a company that had health insurance and offices, company trucks and we were competing against a bunch of really small contractors without any cost structure using borrowed equipment, and then the program also became kind of our number one competitor for customers. And so then we were spending large amounts of money acquiring customers and ultimately competing for us against who the folks that were historically our customers. And then all of that combined with the fact that we ended up hiring at least two additional people and we could no longer deliver reports on site as we had always done because of the amount of paperwork we needed to process. So a lot of struggles that really slowed us down and taught me a lot of lessons in the process. The remainder of my career.. and really that's what we've seen, you know the home energy efficiency business even after billions of dollars going into it over the last five years really struggle getting traction. Most large companies, but companies like Wellhome which was part of Masco, Sears, this long list of Next Step Living most recently, companies that have tried to go to scale in home efficiency have really struggled. So the solutions that we're now seeing come to the table are actually really built on my experience, personally that I had in the market, and the experience of many many others to really change the name of the game and then engage marketplace that rewards innovation and results and kind of deregulates the core business model so that you can see the type of innovation that led to solar. You look at what's actually delivers solutions like SolarCity and SunRun and Tesla, and Nest. These aren't the results of regulatory systems or white papers by consultants. You know its the brutal selection in the marketplace that results in this sort of innovation, and that's really what we want to bring to energy efficiency.

Chris Nelder: So you've actually been about efficiency all the way along, trying to figure out how to get markets to properly value and accept efficiency measures on at least an equal footing to energy that gets consumed. And you've had to create some new ways of thinking about, and doing things along the way because efficiency sort of always seems to be the red headed stepchild of energy transition when in fact it ought to be the first thing we do, long before we think about putting up solar panels or whatever. So your latest move now is to try to change the way that efficiency projects are financed by moving away from a programs based approach that has always been used to support efficiency measures and toward a pay per performance model. So let's start there, what's wrong with the programs based approach that we've been using?

Matt Golden: Well at its core in energy efficiency really is you're stating different than almost anywhere else I can put my thumb on, and you step on a trap in these programs and the trap is really simple. We have software tools like Recurve but many others, you know they could be everything from a sheet of paper with a deemed average incentive predicted savings to very complicated fancy software. They're all making projections. And efficiency is different than almost any other market, we take these engineering calculations and project what is going to happen over some series of years based on whatever the methods are. And then we pay people out. We give somebody a rebate based on that projection and while that sounds really good and simple, it's actually a trap, because what we do when we pay upfront based on an engineering cacluation which again would be underwriting in any other industry is we misalign the incentives. And so if we pay a bunch of contractors out, and they get paid, and the customer gets their incentive up front, we really left them no incentive to care about energy savings, and no ability to profit from projects that actually work better. In fact its the opposite. If you put the extra time and money into making a project that actually delivers. And it does take more time and money than it does to put in the lowest cost box in the run of the mill way the industry does in the market and you don't get get any reward for that work. And so the basic underlying idea behind pay per performance is rather than constrain the industry and basically be forced to regulate because once you've paid up front the only tool that utilities and programs have then to insure and say they can there is create a bunch of rules and go try to make folks do stuff that ultimately loses them money, that's really hard to do, and the result of that is that for efficiency across the country, we have 100 percent program overhead, which means for every dollar that makes it through to a customer we spend another dollar on program overhead to make sure it was spent wisely.

Chris Nelder: Because we're basing everything on projections, and then having to try to go verify that the projection worked.

Matt Golden: Yeah, and make rules to try to again force folks to spend extra time doing stuff that they don't get paid for, and it's really hard to do. In fact I might even say it's impossible to do based on results we have at this point.

Chris Nelder: So how does pay-perpperformance actually work? Like how do you verify that it worked?

Matt Golden: So you may have noticed that pretty much every building has this thing on the side of the building, its called the meter. And we've just got through spending untold billions of dollars putting in these new things that get people up in Marin where I live cancer it turns out, but everywhere else they're fine, called smart meters. That was a joke by the way. But we've invested tremendous amounts of money in smart metering infrastructure that gives us interval data and access to that data remotely in ways that we never had before. So built on that access to data, we have the ability to track energy savings, and not just savings as baseload as we've thought about it in the past, but also savings in terms of time and location using smart meter data. That fundamentally makes it look like capacity. And you know what we can talk more about how pay-per-performance works and why using smart data to measure results in this quantifiable way makes it something that procurement folks and regulators can actually rely on and start thinking about as something they can procure. But the basic notion of pay-per-performance you know what it does in the marketplace is essentially reward those solutions that work. And because we're paying only for results that actually happen when they happen over time it doesn't require the kind of regulation that a program that pays up front based on an energy, on some sort of model or prediction requires to get the results, because we're only paying for the results that actually happen.

Chris Nelder: Yeah I agree. I get that. But I mean a customer's use of energy can change for any number of reasons. I mean so how can a paid performance approach actually distinguish between behavioral changes and efficiency improvements and really guarantee that an efficiency measure was what produced the energy demand reduction?

Matt Golden: Ah, you want to enter the rabbit hole, that was the question, got it. So..

Chris Nelder: Hey man this show is all about rabbit holes, it's all we do here.

Matt Golden: Well the good news is that efficiency is even more fun, its like a rabbit hole and you take an onion and you drop it into the rabbit hole. That's how I like to describe energy efficiency. Someone who just started came up to me after a talk, and said you know I just started, and you know so much more than me, and I hate to tell you that after a decade I actually know that I know less now than when I was in it... Yeah, I'm now aware of how much I don't know. ... How do we make this work? So first you know we've been taking a basically engineering based approach to energy efficiency which is if we just keep modeling, if we just add another air film to that layer in the wall to calculate what heat loss will be, we'll finally be right. You know if we just get a little bit more accurate measurement and a little bit more accurate blower or.. and you're never right. Because these pesky humans that use our buildings and weather changes that are local weather patterns and I bought a new TV and there's just so many of these non-routine adjustments that being there's basically diminishing returns on trying to be right on individual assets based on an engineering approach. So the meter takes more of a financial approach to it, we take a portfolio approach and it turns out when you look at one house and you say you're going to save 20 percent... we have an overall tendency to overpredict almost all the time. And even if we are right on average we see a lot of winners and losers. If you step back and you look at 100 houses that are largely the same, what you realize is while you'd be hard pressed to place a single bet on any one of those houses, you get a really consistent standard deviation meaning you have a very probable, the amount of yield is very consistent. And how many winners and losers you have is very consistent as well. So it starts to look like a car loan. Rather than try to go get a car from the dealership where they don't spend a month trying to figure out if you're going to default, they look at your FICO score and a few other things, and they say great, drop you in the bucket, here's your rate. They don't know if you're going to default but they do know the percentage of defaults for that FICO score. And that's really how we're able to have high confidence is that at an aggregate level, and the pay-for-performance is not directly paid to the customer, its actually paying through aggregators who are market participants, we can talk about who they will be more and more detail but they could be vertically integrated contractors, SolarCity they could be PACE providers, finance companies, they could be any number of players including current program implementeres who play that role but really anybody who can aggregate a portfolio. And it turns out again at a portfolio level and this is really critical for efficiency. This is a very very probable resource that you can really count on when you get enough of these assets in a bucket. And then in terms of your notion of like attribution really, and this is really interesting because California in some recent legislation AB-802 in particular really changed the definition of energy efficiency. So in former programs, where we give out rebates we're really thinking about measuring the use of that rebate like its a coupon, like it's marketing. And so you know if you send out a 20% coupon to somebody in Bed Bath and Beyond and they were already headed out the door to buy bedsheets. You think yourself wow, you just wasted 20 percent, and that's bad. If you're in resource acquisition however you count all savings, and you address the incremental attribution issue through pricing and through markets and you basically count everything. There's a price for a bushel of corn, you check to see if it's in fact a bushel. You know you paid a clearing price, and that clearing price is determined through supply and demand of the market. And so AB-802 actually specifically says we're going to be counting all normalized metered savings, we're going to include code, we're not going to try to only include above code which is almost impossible in reality to quantify. We're also going to include behavior. So we meter buildings if you train your crews to install stuff correctly you get paid, if you train your homeowners to close the windows when they run their air conditioner you also get paid. It's looking at efficiency as a resource for the grid. And that's the right way to think about it as well. If you're in a procurement department of a utility you don't care why the savings were there, you care that a particular set of buildings are demanding fewer electrons than your forecast would have said otherwise. And so you don't need to buy more production infrastructure.

Chris Nelder: OK. So I think the answer to my question was you do disappear the margin of error down basically a larger dataset.

Matt Golden: Yeah. That's the law of large numbers washes out the variance. And you count all of it because it's a resource, not a coupon.

Chris Nelder: So you count all of it and then you're able to basically derive what a single customer's portion of that would be.

Matt Golden: So this is where the market comes in. The opening meter in California is called the Caltrex system. The core of a marketplace is sellers agree on a price for a product. To make this all happen, the first thing you've got to do is establish what is energy efficiency which sounds like a silly question but we've never had agreement on this point. And this is particularly hard to measure because you can't actually measure efficiency, its a calculated value, because its the difference between a counter-factual, what would have been if you hadn't of retrofitted building and what is in fact now their bills. So you have calculate that counter-factual. So what enables this whole system is we run the opening meter on a process that PG&E has been leading that I've been working with PG&E for the last three or four years on is getting all the powers that be in California and soon to be other states as well, but getting everybody to agree on how you calculate a unit of energy savings in a uniform way which makes it reliable and consistent and also makes it not just something a utility can do something that everybody in the market can also do. So if you're a PACE provider, or an aggregator of any sort and you're going to get paid over time now, not up front rather than get a rebate you're now getting a cash flow, that cash flow is a result of being competent in your yields times the price with an investment grade counterparty called the utility which turns efficiency into a long term cash flow instead of an up front payment. And so if you can have confidence in that cash flow as an aggregator you have to have this aggreement on how this calculation is done. The problem we have right now is if you gave the same hundred homes to five different engineers you probably end up with seven answers. This is the punch line but it's true. They'd all be right. And that's the problem. So we the opening meter Caltrek process is getting an agreement on what exactly is a negawatt and how its calculated. And then because of things like Share-my-Data and PGE Territory which is our version of Green Button which gives us access to AMI data, we can talk about the smart meters again outside of the utility firewalls with customer permission and the fact that we've standardized the data being used from the state of California on something called Performance XML third party aggregators have all the data they need not only to track savings in their portfolio using an API with the utility, getting load shape and gross savings but also the data they need to start to figure out what works, what sets of energy conservation measures and what types of houses deliver what types of load shape. And that all becomes a function of the marketplace.

Chris Nelder: OK so the cash flow obviously originates with the customer, they're the ones who are actually putting the money out, they're spending essentially what they were before the energy efficiency measure was made and then there's the delta for what's actually being consumed in energy and that becomes the financable cash flow?

Matt Golden: That's a great point. So there are actually three cash flows in most energy efficiency transactions. There is the savings that the customer experiences on their bill. And in this model they're going still receive those savings on their bill every month. However you can securitize those savings as well, that's basically an energy services contract, is aggregating and securitizing bill savings, so that's cash flow number one, savings on your bill. Cash flow number two is some sort of straight finance, we often call it efficiency finance but its underwriting generally credit or asset or the customer or potentially being underwritten by a PACE assessment or on-bill financing mechanisms, but we call efficiency finance but we're not underwriting efficiency really it's not efficiency finance, it's just finance. So that's the next cash flow and that's you know finance provider says we're going to provide the up front capital. You've got a 700 FICO score, you're going to pay us back. Right. That's cash flow two. We're adding a third cash flow to the model which has its origins that are similar to a rebate form of thinking, but it's basically saying to the utility, because we're going to go retrofit an aggregator said this, we're going to go in Marin, because this happens to be where I'm standing, you know we have a lot of solar production which is driving the cost of energy at 2 pm into negative territory, where holy cow 4 pm is a huge problem on the neck of the duck, where the duck curve comes rocketing back. We're going to invest in energy efficiency load shape, permanent load shape change to reduce the neck of the duck and that's going to mean we don't have to build those transformers anymore, we can buy half as much storage as we would otherwise going to have to buy. And that's the capacity value. And that's the third cash flow, and that's actually what we're talking about securitizing and aggregating in this model. It's not individual customers, it's actually the customer signing that value over to the aggregator because again there's a real moral hazard on individual customers, you know you don't want to be sending people checks because they went to Mexico for the summer. And you don't want other people who get a bum retrofit which does happen, to not see the savings and not get the value of their capacity because they had a bad contract. You know so we aggregate it at the portfolio level and that's how we have companies who have the data to understand what works, to basically translate that long term cash flow into better products for customers and contractors that maximize that benefit, that savings benefits the utility. And so you know in a competitive market that cash flow it's not like they get wait around to get paid, to get paid over time. They take that cash flow which is very reliable, because again portfolio level yields are consistent, you got a utility paying the other side of the transaction, you know that becomes something you finance, its going to become structured finance. And the good news is, is that type of structured finance, project finance, financing cash flows from projects that's actually what project finance needs. Project finance is not financing projects, its financing the results of those projects, the cash flows and that starts to look like how we build energy infrastructure. If I was to go to a large utility in say I'm going to build you guys a natural gas plant, or and I go to my investors, I have a PPA with the utility to buy the power I'm going to be generating, I'm going to have to have good credit and that's going to have to be a strong asset but before I get a billion dollars to build my plant or whatever it's going to cost that investment is underwritten by the fact that I'm going to generate electricity and sell it. And that cash flow is what underwrites that investment. So what's really interesting about this model is once you kind of shed all the layers of energy efficiency that we've become very accustomed to, the rebates, and the programs, and regulation valuations and all these sorts of things, you look at it, and we end up with something else that looks just like every other market, we've agreed on how we measure efficiency as a drive value we've agreed on the calculation, utilities send a price signal, they say you know we a certain amount of demand capacity in certain periods of time, you can bid into a market, this is a little bit future at this point, but this is where we're going. And it will arrive at a price based on supply and demand, and the market will figure that out. And you know as long as it's cheaper than alternatives is it ends up being a good deal. And what you end up with is long term cash flows by all these aggregators that can be securitized and brought into the secondary market, just like car loans or anything else. You're local car dealed doesn't sit on your car loan, it gets packaged up and brought into the marketplace and that's how capital gets injected into the market.

Chris Nelder: So does the initial capital to do all these efficiency measures, is this still provided by the customer or is another entity here that's stepping in to be the financier?

Matt Golden: So there isn't one answer which is a little bit of a frustrating answer, but that is honestly the truth. One of the beauties of this model that makes me confident in it is that we don't have to fortell what the winning business models will look like if that makes sense, the market will, and there's probably going to be a lot of different answers that work out. But in general we've found that customers don't want to wait around and get paid, and somebody does have to pay. So I think what you'll find in the market most part is you know, let's just take PACE as an example, the residential PACE in California is the most exciting thing that's happened in energy efficiency period, since I've been involved. They did $500 millin dollars in residential efficiency projects just last year, entirely with private capital without using incentives.

Chris Nelder: We should probably explain for those who might not have heard of it that PACE is property assessed clean energy, and why don't you tell them how it works?

Matt Golden: Yes so its fairly straightforward honestly, primarily rather than looking at customer credit, what PACE does is it assigns, it creates an assessment through your local government whereby you pay back your PACE assessment which can be used to pay for renewables, water efficiency and of course energy efficiency that are affixed to the house based on an assessment on your tax bill. And the beauty from a financing perspective os that the tax man always gets paid. And this means that it can survive foreclosure, it can stay with your house when you sell it, and you can have basically awful credit as long as you've got a house that the taxes are up to date. That has a certain amount of equity that varies a little bit from place to place but you have equity value. So this eliminates a ton of barriers. And it makes for a very low cost long term transaction. And so you get a relatively low interest rate for what it is really not a secured product for up to 20 years, and even longer than that for some measures like solar that actually have a longer life span. And what you have in the PACE environment is a bunch of competing very well funded companies that again have gone from next to nothing four years ago to a half a billion dollars just in California last year and growing at a just an alarming number, but an amazing clip, just absolutely scaling. The downside to PACE, I wouldn't put this as a problem for PACE, I'd actually think is a policy problem, but we've given the PACE providers no reason to care about energy efficiency. They get paid based on moving money into the market full stop. And they care about energy efficiency but they're businesses you know and their investors want a return on their investment and they're not going to take one for the team. So let me go back to your question, what does a customer expect, again they're going to have a lot of choice. That's the first thing. There's not going to be one program, contractors and customers are going to be able to choose those business models that actually work for them. But I think for the most part you'll see there's PACE assessments in the market right now that are getting a lot of traction. The future, it may be as simple as when the HVAC guy shows up they're not going to just sell you a furnace anymore, they are going to sell you a furnace with some duct ceiling and insulation, replace a few incandescents with LEDs at the same time and put a smart thermostat on the wall and the whole thing is just a lot less expensive than it would cost you just to get the furnace in the first place because they are maximizing the value of that cash flow on the back end and they are using it to cost up front of making a more valuable product, and beat the competition.

Chris Nelder: So is it like buying a car where the customer can just walk in and basically buy the car along with a financing package from whatever bank is providing it and then walk out after making their first payment in a down or something like that?

Matt Golden: Yeah I mean those models definitely exist in case law. I don't know because you'd be out of pocket but the only difference is that efficiency today and we hope this will spur change. Really but is more so than it is. So it's probably a provider, it's more like generally a customer having either something that's broken or some sort of problem and then you know enlisting the help of somebody who will sort out basically what could help actually solve that problem, and how to finance it, so its more sold than bought and yes that's basically what it looks like. So to a customer everything I'm describing it sounds very complicated frankly, if you're a contractor or a customer the car analogy is just an easy one but every time, if you go buy a cell phone or a car or any of these things you're actually interacting with the system that's almost exactly like I'm describing. It's just that it's totally invisible to you as a customer. And it's also invisible to you as a... if you work in a car dealership you don't really understand what's going on to manage risk on a portfolio of car loans. So from a customer perspective you're just getting a better deal. And there's more choice in the marketplace is what it comes down, but one of the real strengths around pay for performance is just like when the solar industry started, nobody had a master plan around leases and PPAs and how these companies go to market. It was really that brutal selection, the fistfight that's a competitive market that resulted in those solutions that work for customers and also made money and work for capital markets. And frankly that's where the solutions are going to come from in this approach and that's it's strength is that you know I have my opinions on what I think will work in the market and so do you perhaps and many people do. But when you change this model and say we're going to track and pay for results we really can allow the market to function and let customers choose the business models that work best for them. And same for the contractors. You know let them choose the models that work for their customers and they make money at. And rather than being on four year program cycles with two hundred page report at the end to tell you how you did, we'll be on the cycles you get in private sector which is generally a couple weeks in a payroll cycle where you don't change your debt. That's the real advantage the private sector has, it's not that private companies are a lot smarter, frankly it's probably the opposite. It's the feedback loop and the fact the continuous improvement, and the fact that if your ideas are any good you clear out pretty quick which is a market efficiency.

Chris Nelder: Right, so on March 25th PG&E filed a plan with the CPUC to launch a residential pay for performance program as part of what's called the High Opportunity Projects and Programs or HOPPS, the program will not only take a pay for performance approach to residential retrofits, it also opens the door to aggregators who might manage a portfolio of such projects as you were discussing a few minutes ago and get paid by PG&E for maximizing the energy savings from that portfolio. So why is this new PG&E plan interesting and what's the advantage of having aggregators playing a role in it?

Matt Golden: Well I think this is interesting. I mean almost all for the reasons that I just described. There's been other attempts at pieces of this basic model but I think this is the, PG&E is pulling together all of the different pieces necessary to establish what I would consider the first true energy efficient market or at least the beginnings of the first true energy efficiency market. So I think that's why it's really exciting, and it is an early pilot so it's basically a simple buy it as one can make it because we are introducing a whole bunch of new moving parts into the model all at once. But in that pilot as well, PG&E references that this is really about creating a framework in the beginning and moving towards you know energy efficiency as a procurable resource. These are really just the first steps. That's what makes it so exciting is that frankly it's not just another program. It really is with intention that this be starting to build the systems and start to kind of create the structure of a market that can lead to procurement of efficiencies capacity over the next few years.

Chris Nelder: So what is the role of aggregators?

Matt Golden: The aggregators are really the businesses, and they can be PACE providers, they can be vertically integrated contractors. Some of the largest program lenders in the country are also looking at this model. That's the engine of innovation right. Those are the companies that are competing to use the data that's now available and to put together all the pieces that one needs to have a consumer offering. The upfront financing you know, having the data to figure out what types of energy conservation measures work best in certain houses, how to package that up and train contractors to deploy, all the different facets that are necessary to get that really simple customer offering into people's hands, and so kind of critical was models, you know again, we're not writing checks to individual homeowners. It's the aggregator that's assembling a portfolio of energy efficiency projects and entering into a contract and then selling thos savings from that portfolio to the utility, it's not a traditional utility. And it's the alignment of interest, so really you know if you take those companies that are competing with each other in the market, it means that those that can figure out how to deploy energy efficiency projects that deliver in the case of the pilot the most savings but ultimately the most valuable savings will make the most money and be the most profitable. So it sends the right market signal, and then we want the money to flow through industry, you know and because that's actually, at the beginning we talked about how paying rebates upfront misaligns incentives, right you get paid in advance. You have no reason to care about long term energy savings. It's similar to solar with a solar lease or PPA, where performance risk moves into the market. Now these aggregators and the contractors and ultimately the secondary markets that are going to be investing in the proceeds from these transactions, and in the long term cash flows are at risk to the performance, and that's really what it enables us to deregulate. And there's a great analogy for that in the solar industry where you can see that if you want to securitize residential solar projects, things like quality assurance are a flat out requirement because as you know those products aren't just straight loans. They have a performance guarantee associated with them, so now all of a sudden rating agencies in senior capital markets care about the quality of the underlying projects because they are actually at risk, so kind of the one of the more key things in addition to just bringing capital into the marketplace in a way that'll make contractors more profitable is creating that alignment which enables us to step out of the rebate trap and regulators don't have to regulate every step in the process.

Chris Nelder: So you actually see like fixed income markets or maybe even yieldcos getting involved in this space?

Matt Golden: Yes. Not tomorrow.

Chris Nelder: Yeah. Well it's going to take a while to build up an adequate portfolio that would be the right deal size for them to be interested in that sort of thing anyway.

Matt Golden: Yes all that's true. And frankly we just need more data, it's all a function of the data, and we have a lot of data now but markets will get involved based on the data we have today but they really need more first hand experience to gain enough comfort. But I can tell you that over the last decade I've had many meetings with our friends in the capital markets working for Goldman Sachs, and Deutche Bank and all these different players and you know usually when we talk about efficiency they look at me a little cross-eyed to be honest, they can't quite figure out what I'm talking about historically.

Chris Nelder: Well no one's turned it into a good cash flow yet.

Matt Golden: Right, but as we turned this corner and actually I'm not going to try to take all the credit, they've been hearing from people like Richard Kaufman in New York. And I've sat down with them subsequent to that and that you know, they don't know how to trust me from Adam, but they know Richard Kaufman, so they say he's talking about these markets and we start to put these pieces together and I've had these kind of epiphany moments where they say this is just a structured finance. You know if the yields are really as stable as you're showing us you know when you have an investment grade counterparty those are deals we do all day long. You're going to have to get it rated, but if you can get it rated, and also if you can get us 100 million dollars worth it. That's the other piece, those are the kind of deals we can do. And then they start talking about wow, well bring it to use before the traders see it, that kind of thing. But you know we're not there yet but it makes sense to them, one of my favorite quotes from Cisco Devries the CEO of Renew Financial and actually really the guy that came up with PACE to a large degree, they've done some securitization mostly actually of unsecured long products. But everyone says, how did you manage to do that? Everyone has talked about doing that for all of these years. How did you do it? His explanation is really simple. We didn't go to Wall Street and try to explain to them why efficiency is special. We put data together, we made efficiency transactions essentially look like a mortgage backed securities transaction, we made it look exactly like everything else.

Chris Nelder: Exactly. So I've heard you talk about demand capacity as a distributed energy resource. This is kind of a curious little twist on the usual meanings of these words so what sort of demand capacity are you talking about and how is it a DER?

Matt Golden: Yes, so I felt like we needed a new name because there was not enough acryonyms in efficiency. Well actually I think it deserves a new name because efficiency tends to mean baseload deemed savings programs and energy in you know physics models. And so the fundamental difference between efficiency and demand capacity in including location and time dependencies. You have to know when and where it happened. That's really critical if you are a grid planner. I'm going to save energy in PG&E territory they say that's really interesting, go do that. But to make it something they can actually use, and think about as part of their actual physical energy mix to make sure the lights stay on, they need to know when and where it's happening. And again that's newly available information to us. We're still learning a lot about honestly, because we can now be interval data, and we can get it out of the IOUs and it's still very basically impossible to do an individual assess. But when you start to get large enough n's and big enough numbers you could end up with pretty high confidence intervals in the load shape that you're generating, and so you know basically this when you squint and back up a little bit, it makes efficiency look like other forms of capacity comes, is what it down to. And so we kind of coined the term demand capacity because I can't say I love the term frankly but it needs a new word.

Chris Nelder: So you're trying to integrate the concept of locational marginal pricing in with efficiency basically?

Matt Golden: Exactly, and just trying to draw a distinction between the two because it is really a form of demand response that's passive demand response that's permanent load change.

Chris Nelder: Right. That's the part I was struggling with there because when I think of demand response I think of something that's sort of transient that just happens for minutes or hours or whatever. Not something that's permanent.

Matt Golden: That's a different resource that can also be, they are all attributes of the same product that can be sold in the different markets, there are different dimensions.

Chris Nelder: The duration you mean.

Matt Golden: Yeah there's permanent load shape change, that's one attribute that can be sold, there is dispatchable demand response which is what we typically think about, there's carbon value certainly depending on how you're regulated. In California its already rolled into a price but elsewhere that can be disaggregated. There's even efforts to monetize health values that are real and so all of these are different attributes that can be actually monetized very much in a very similar way over time.

Chris Nelder: Yeah. So I'm glad you brought that up because that's what I wanted to talk about next. I mean it's clear enough that you could pay for energy efficiency retrofits based on actual performance but what isn't straightforward is how much we should pay for it, how it should be valued because clearly if you can perform demand response with it or avoid investment in new distribution capacity because you're reducing demand then you should be able to get paid for more than just the kilowatt hours or btus of energy that are being saved. So what's the state of the art in valuing energy efficiency?

Matt Golden: So first of all we have all these smart people trying to develop business models. But answering that question is what is the value for efficiency, that's the elephant in the room. That's what we should be spending our time on honestly. You know we're just figuring that out for solar and we are just starting to even figure that out for storage. These are very complicated questions. So there's a couple of pieces to it. In the pilot that you referenced from PG&E we're starting out really simple. So we're saying we should set a price as a discount rate to the current total cost of the program. We're spending x per kilowatt hour, when you add in incentives. Let's discount that 25 percent. Wouldn't that be great. We can get more savings for 25 percent less cost. So to start with its really sythentic, we're just picking a number because no one knows how to value it to be honest. But as we move down the road there is a broader conversation to really think about all the attributes to make sure it's fairly valued. But so two things, one is that rather than be as cheap as we can make it which is efficiency, you know everybody wants the two and a half cent kilowatt hour. You know we're sending out coupons, so the goal is how cheap can we make those coupons. Right. But we should now be thinking about energy efficiency really as competing against the levelized marginal cost of the alternatives in the marketplace. And if you were going to have to buy batteries at 18 cents a kilowatt hour then I can go in and do fancy heat pumps or whatever the technology is, or insulation at 16 cents a kilowatt hour and get you the savings in the right place at the right time in the real, that sounds like the price. It's cheaper than the alternatives it's a good deal, it's a different way to think about it. And the most important thing is we need to introduce supply and demand. And honestly this is not a good podcast to dive into to a two hour conversation.

Chris Nelder: Oh you'd be surprised. Maybe next time.

Matt Golden: So you know there's a lot of concern about, because efficiency as it has been has all been about incremental savings. We only want to pay for new stuff. There's a lot of worry about free ridership. This notion of paying for stuff that would happen anyways. And that is also largely solved by markets. And so if we're trying to procure a limited amount of capacity in a certain area, say we need two negawatts worth of savings, and you have competition to get into that procurement, that will set the price right. And if there's really much cheap or free savings that would have happened anyways, that means you have a lot of supply and the price should fall. So there's a lot we can do you know within that framework is to allow competitive markets to arrive at the price and as long that price is, you can rely on it and believe it's real and you're really not going to build those transformers that you would have otherwise. And it's actually less, then it's a good deal. At the very beginning when I started my first job out of technology into renewables and energy efficiency because the company I worked for sold a megawatt at $9 a watt, and got a $4.50 rebate. Never going to happen again in the history of solar for sure. Right. One should not believe markets are perfect because they're not, they have to be regulated. So there's a big task there. But I do believe that markets are fairly perfect at one thing which is in a transparent competitive market all the fat get's squeezed out quick. And so I think we do need to kind of trust that these markets will arrive at a price that is reasonable if you allow them to emerge and be competitive.

Chris Nelder: I guess the efficacy of this market will really depend also on the players like whether the right entities are participating and what they're buying and selling. We've talked on this show in previous episodes about some of the components of valuation that are being explored for things like ancillary services provided by devices that can do demand response. We're valuing the various services that can be provided by storage systems. If the valuations were unbundled so how do we make sure that the full suite of services that energy efficiency can provide get traded in these markets and not just left out of the market as they are today?

Matt Golden: We're not, in fact we also can't, we don't know where it comes from at the level that we're measuring. We can actually tell the difference between behavior, we might be able to in the data disaggregate it out but honestly we don't care and persistence is handled because we're paying over time. You know we don't have to worry about figuring out if we think it's a reliable piece of equipment, we're paying for the savings as it actually happens. So it's pretty much agnostic to technology. If you have some new home energy internet of things device you don't have to prove to the energy commission that it works, take your own bet. You know if it works you make money, and if it doesn't you don't. So it's really agnostic to technologies all around. And then you know within market design you can do things to encourage the type of things that we want to see happen in the market. Some of those things may not be about being the cheapest savings, we might want to encourage deeper retrofits. You can do that by how we establish markets, we can do procurements that require certain depth of savings on average, we could even do things aimed at certain technologies, as long as we recognize that we're doing those things not in the name of cheapest energy, and we should be held at the same cost affecting this thinking but to achieve other policy goals, that would also include helping low income families for example. Right. That's in my world perfectly acceptable, we just have to acknowledge what we're doing, its not about lowest cost resource any more.

Chris Nelder: Right. So among the many projects that you seem to somehow be involved in, tell me about you're also the director of the Investor Confidence Project. What's that about?

Matt Golden: So ICP and Oakland Energy Efficiency are very similar beasts. So at a macro level, the way we're approaching residential energy efficiency and also smalle medium business which is another really good approach to take a data driven approach which is really the pay per performance approach that we're talking about, so if you have a large number of similar things, put them all in a bucket like we're doing, and that works really well, the Investor Confidence Project which is an Environmental Defense Fund project that I've been running for the last four years is really trying to put the foundation in place to achieve a number of goals in commercial and multi-family that ultimately arrive at the same place. So what we've done with the Investor Confidence Project is taken commercial multifamily projects wich are really all over the map. Each one is unique assests is what it comes down to, even if they're not totally unique you'd be hard pressed as a building owner or a utility or any investor to know what you're looking at without really expensive due diligence because when we have multiple standards in the marketplace that you can apply to the same project in many different ways so you end up with a very heterogeneous pool of projects. So what ICP does is we certify what we call investor ready energy efficiency projects. This is an underwriting certification that says that a given retrofit project has followed the appropriate standards applied in the right way and then documented in a standard way across its lifespan from baselining and savings calculation and then having it commissioning O&M and measurement verification plan in place that can executed on during the performance period and then its certified by a third party, an independent review by an engineering firm. And that's basically creates the confidence for a utility program or a building owner, or an investor that they are looking at a standard asset. And in the near term this reduces underwriting costs and technical due diligence where often the same project is going through the same 10 plus thousand dollars review by three or four different actors. So it creates a lot more liquidity and lower transaction costs so you can get more deals done, but it also establishes an asset class, a standard bucket of projects that you can start to aggregate together into portfolios and then underneath that we have documentation we're actually mapping that and about to release a data specification. So ultimately what you end up is classes of standard projects, the underlying data that you need, the analytics to understand the yield. So the Investor Confidence Project right now, we actually PG&E announcned two pilots last week, one was pay-for-performance in residential. The other was an on-build financing pilot using the Investor Confidence Project certification. We're also working with New Jersey on their pay for performance program, we're about to be announcing some pilots in New York, we're working in Texas with their PACE program. We're also in about six countries in Europe, througha European commission as well. And so if you think about the Investor Confidence Project it's actually kind of building the foundational elements so that as we get increased scale and larger volumes in commercial and multifamily, very much the same approach can come into play.

Chris Nelder: Gotcha. Wow fascinating stuff. You certainly have done a lot of work my friend to make all this stuff work I'm just so impressed.

Matt Golden: It is a lot of work. We still have to make it work though. So that's the exciting part is a friend of mine who's been involved, you know saying, ah you should be so happy. I did a momentary celebration when PG&E submitted their pilot last week. But honestly all of this has been building the additional elements in place and now we actually have to go do all the real work.

Chris Nelder: That is truth right there. Wow. Well that was a quick hour. Thanks a lot Matt I really appreciate you taking the time to be on the show and share with our listeners how all this crazy stuff works.

Matt Golden: My pleasure. I'm sure you'll have some way to share some of these links as well.

Chris Nelder: Absolutely. I have all sorts of homework for my listeners in the show notes.

Matt Golden: And my advice by the way is if you've been in efficiency and a lot of this stuff sounds a little daunting. Stick with it. And this is just my prediction but I had this personal transformation myself. I work with a lot of folks who come out of efficiency and we get very comfortable with way we have been doing things. And so just trust me and put in the time because once you're able to snap out of the paradigm and start to really understand how this approach works while it looks very complicated on the outset, it actually becomes a lot easier, and you find that this basic approach to efficiency, it's actually a lot less brain damage than what we've been trying to do.

Chris Nelder: And that's the upside of the whole thing right there. Yeah. Cool. Well thanks a lot Matt for being on the show.

Matt Golden: Thank you Chris.