Competing With Data & Analytics
What to Read Next
As vice president of engineering for the energy management company EnerNOC, Hugh Scandrett is deep in the trenches of bringing energy management applications for the smart grid to a wide variety of companies — utilities as well as commercial, institutional, and industrial customers.
Scandrett joined the Boston-based company in 2011, and as head of engineering he leads the company’s software development, engineering, quality assurance, and R&D teams. EnerNOC supplies guaranteed electrical capacity to commercial end-users, utilities, and wholesale energy suppliers, and accompanies that capacity with energy intelligence software. It helps companies figure out how they buy energy, when they’re seeing the highest demand, and how to optimize their usage.
In a conversation with Sam Ransbotham, an associate professor of information systems at the Carroll School of Management at Boston College and the MIT Sloan Management Review guest editor for the Data and Analytics Big Idea Initiative, Scandrett says that being able to predict energy usage is key to saving money, and explains why turning on the AC at 7:00 a.m. on a hot day can be more efficient than waiting until 8:00 a.m.
From a data perspective, what is important to EnerNOC?
The key value EnerNOC delivers to its customer is the technology and best practices to help companies become more energy consciousness, energy efficient, and essentially save money and improve operations.
We do that by focusing on three key areas: The first is how you buy energy, the second is how much you use, and the third is when you use it. Those three things come together to make a big difference in how much you spend for your business’s energy.
In the past I was a chemical engineer, and there was a guy at a plant in the Tennessee Valley concentrated on peak electricity. He focused on that meter and made sure that we didn’t go over the past peak.
Yes, that’s part of “when you use it.” You’re billed not just for how much you use overall, but how much is the maximum stress you put on the grid. Going over peak typically incurs what’s called a demand charge. In those circumstances, it might actually cost less to use more kilowatt hours, for example by starting some equipment earlier and avoiding that big spike in demand.
I’m sure that’s changed since then.
No. No, it has not changed. In fact, some companies have as much as 30 or 40% of their cost related just to the charges associated with their peak usage.
The peak that they hit also determines essentially the size of the pipe that needs to be provisioned to serve them. That high watermark has a huge implication on what companies get charged, still today.
I guess I was hoping that it isn’t a person watching the needle anymore. I’m hoping that at least that has changed with the Internet of Things and ease of getting data from lots of devices.
Oh, that has changed. We have technology, both physical meters as well as APIs [application program interfaces] and ways to collect data from utilities that don’t require direct metering. So we can collect hundreds of thousands of data streams that tell us what consumption is going on in real time, and can account for the impact of different things, like weather, on overall demand.
If analyzed and presented effectively, that data allows people to make decisions about how to change their operations, such as moving production runs or staging equipment start times.
So our data lets us look at both simple stuff and really complex stuff.
Tell us about some of the complex data analysis you do and how that helps companies save energy.
One of the big issues is being able to predict future demand. Many of our customers would love to adjust operating schedules in order to avoid peak demand charges, but running around in fire drill mode isn’t always feasible or good for operations. But, if you can predict consumption and you can translate that into a savings opportunity in an amount of time that people can react, that’s a game changer.
So we predict a company’s usage based on analytics that look at weather, degree of sun azimuth, and a whole set of other parameters, and as a result, we can tell you that tomorrow at 2:00 in the afternoon what your peak is likely to be, within a few percent, adjusted for weather, adjusted for your tariff on that building, based on all the history of it — and we can map that information to your specific utility tariff (aka, the way that you get billed) and tell you specifically the costs that peak will incur.
We then can provide techniques for minimizing peak usage, like pre-cooling a building. We notice that you have a startup at 8:00 in the morning. If you started that at 7:00, it’s a little less expensive, gets the building cooler, and it turns out you can ride through the morning and then do the same thing through noon, and you can ride through the afternoon peaks, because you have done cooling management: looking a little bit ahead. And just that can save as much as 10 or 20% of your daily costs. It’s remarkable.
What strikes me as hard is predicting the peak demand usage tomorrow. Is that a fair assessment?
Yeah. Think back to the chap who might have been watching the meter 20 years ago to see that it didn’t go over it. The electrical energy industry is so ripe for technology to help it. And the intersection of software IT, analytics, Internet of Things, all the various pieces — that industry is set to have huge, highly leveraged sets of changes.
For many companies, even simple things, like being able to accrue energy costs during the month so you know how much your bill will be at the end of the month instead of waiting 45 days to get a utility bill, are still really difficult. Bringing that type of information into real time — the same way it is for virtually every other major operating expense — and also translating it into dollars, is huge for many of our customers.
This is not rocket science. It’s basic energy accounting, or accounting for the cost of energy, in semi-real time.
Forty-five days seems like a huge time to have to wait. What’s driving that delay?
That’s utility typical billing cycle. During that time they’re sending paper bills that go in the U.S. post. It arrives in the mailroom, makes its way to some accountants or energy manager, and is looked at against budget. Often times, the people paying the bill have absolutely no visibility into whether or not it’s even accurate. They just pay it.
That’s opposed to, through our application, literally logging in the day after and saying, “What should I be accruing, based on all the consumption that I’ve just had in this building or this portfolio of buildings, with tariff-adjusted in U.S. dollars or in global currency?” Our system does that.
This has very high value for companies that are struggling with energy management. Again, there has been so little technology that is consumer- or business-facing. Even the utilities that spend most of their time on energy generation and transmission — delivering the goods, if you will — haven’t historically been focused on delivering products and services in a way that help the end consumer make better overall decisions, though that too is dramatically changing as utilities look at new software to adapt their business models.
What makes it difficult for people to adopt these approaches? What’s the challenge?
The real challenge is mostly centered organizationally around “who is the person who cares.” It’s really simple for you and me to say, “Of course they would care if they could cut their energy budget by 20%.” There’s no question there’d be lots of people who would see the benefit of saved operational expense.
But most companies aren’t set up to have energy management as a function. There are people who build the buildings and connect up the energy services. There are people who pay the energy bills. There are people who get the calls in commercial real estate and building management saying “this room is too hot” or “the AC is too cold.” But overall there really is organizational “immaturity,” but I don’t mean it negatively. It’s just, there haven’t been tools for people to effectively look at and manage and take action and then see the results.
What you’re describing is a little bit different than a lot of the people we talk to who are shifting from intuition-based decisions to database decision-making. In some sense, what you’re fighting is a third option, which is not even thinking about energy usage at all.
Yes. Just think of commercial real estate. The hot and cold calls are what drive the activity, right? “This room’s too hot. I’ve got a conference room with 400 people.” Are you literally going to walk in and tell them, “No, I’m not turning it down a degree, because it’s going to cost us 30 bucks?” No.
But if you’re in an empty building and you didn’t know that the AC was running full bore in this whole unoccupied section or unoccupied hotel rooms or whatever the thing happened to be, yeah, you’d go and do that. But you have to have visibility. And if you don’t have visibility and you haven’t set up a person whose job is to care about that level of efficiency, then this just falls in the cracks between a number of people’s existing roles.
A lot of what you’re trying to do is provide more visibility while also having the data about whether something is the right decision.
Absolutely. Insight has got to be brought back to a business decision with costs, in real time. That’s the same thing with sustainability. Energy management oftentimes gets lumped into sustainability initiatives, which everybody wants to support, but oftentimes they don’t have quite the closed loop of accountability.
We don’t lose sight of the altruistic part. We’re on a mission to ultimately have very positive effects on the planet. But it isn’t the only part.
Yes, sustainability is a clear benefit, but don’t skip over the data.
So many companies start off with green, and green for good’s sake, and I think people are just tired of hearing about that as the first thing. There’s a lot of good causes in the world. If you can highlight the business relationship and saving money and value, then the green part is a bonus that gets me a little more excited, that gets me to act a little more quickly. But it isn’t foundational to making a decision.