Subscribe to our newsletter
Why Hawaii is - and isn’t - the perfect test bed for solar-plus-storage

Map of Oahu's distributed generation, as of May 2014. Image: HECO.

Hawaii has been namechecked as a ‘test bed’ for how grids can migrate to higher levels of PV penetration so often it has become something of a cliché. With around 300 watts of rooftop PV installed per capita and high fuel import costs, the islands are aiming to go 100% renewable by 2045.

It hasn’t been easy for Hawaii to integrate its high installed PV base. As that cumulative capacity keeps increasing, measures including a temporary stop to net metering applications were recently put in place.

More pro-actively, HECO, Hawaii’s main utility, was recently awarded US$2.4 million of funding from the Sunshot Initiative, the US government’s targeted funding support for commercialising technologies that reduce the cost of solar. HECO matched that funding, and selected a number of companies in the fields of PV monitoring and forecasting, energy storage and other smart distributed generation areas to participate in SHINES (Sustainable and Holistic Integration of Energy Storage and Solar PV), which looks at incorporating more PV and storage onto grid networks.

Stem Inc is a provider of intelligent, software-defined energy storage systems, currently active in the US’ commercial space. The company’s business model allows businesses to reduce their energy bills by shaving peak demand. One of six companies that came together in Hawaii in November to form what is claimed as the first trade group for distributed energy resources in the US, Stem's Hawaii programmes are also being supported by SHINES. Energy Storage News spoke to Stem’s VP for Hawaii operations, Tad Glauthier.

Most trade groups are formed to advocate for or advance one technology in a specific direction, whereas with the Distributed Energy Resources Council (DERC) you’re looking at getting different distributed resources to click together. Can you take us through the reasoning behind that?

Tad Glauthier: We’re trying to look at the issues more from the utility’s perspective. We had a round table with the head of demand response at HECO and he said we think of all those customer-sited resources, things that can reduce demand, that change what it looks like to us at the customer’s meter, they’re all in a similar bucket.

If we’re doing things to reshape the customer’s load profile, or cause their load to change rapidly as a result of some condition on the grid, all these different technologies are just different forms of demand response, to use the old term. They might be really fast, they might be responding to signals in a smarter way, but they’re reducing load at the customer’s site, or increasing load at the customer’s site. That’s what DERC is trying to acknowledge.

Utilities haven’t traditionally been as accommodating of solar as we might hope, but is it true to say that it’s a different proposition with other distributed resources, particularly energy storage, which rather than putting a specific set of demands on the utility like PV, can help the utility and grid?

TG: I think traditionally the utilities have seen these customer-sited resources as presenting additional complexity that they need to now manage around. We want to find a way to turn that equation around so that customer-sited resources and specifically customer financial investment in energy assets can be taken advantage of by the grid as opposed to creating bigger problems for the grid.

The first step in that is giving visibility to what those assets are doing and forward-looking visibility about what those assets are going to be doing. Grid operators are always trying to balance supply and demand a minute from now, a couple of seconds from now or an hour from now and so on. Traditionally, it’s not as if every rooftop PV system sends a forecast to say how much power it’s going to be sending back to the utility.

There’s been a lot of work in Hawaii into getting visibility into PV and there’s one of the best forecasting projects in the world going on here, an irradiance network of sensors on the ground in Oahu, to measure the intensity of light. It then marries that with satellite data so that they can create hourly, six hourly, 24 hours-ahead forecasts to see how much energy these rooftop PV systems will put onto the grid in the future.

We’re also using 29 distributed storage systems on Oahu for the customer’s benefit to lowering customer bills through demand charge reduction, but also aggregating those systems and showing them to the utility as a single resource that they can use very quickly with what we’re calling grid response. We didn’t want to call it demand response, as it’s a much more firm kind of resource than traditional demand response has been.

They’ll see this megawatt of energy storage devices, they’ll see exactly how much is available right now, they can ask for all or part of that to start discharging in response to something going on the grid and they’ll get what would look like a megawatt of lights turning off.  

They’ll see this megawatt of energy storage devices, they’ll see exactly how much is available right now, they can ask for all or part of that to start discharging in response to something going on the grid and they’ll get what would look like a megawatt of lights turning off."

Why Hawaii is - and isn’t - the perfect test bed for solar-plus-storage

Stem's 'intelligent' storage systems for the commercial space combine big data and predictive analytics with energy storage. Image: Stem Inc.

Examining the Hawaiian test bed cliche...

When will meaningful data start to emerge from SHINES? What’s the programme outline?

TG: There are several projects here, PV forecasting into the near future, aggregating customer side storage, and a third effort around smart inverters. Basically SHINES is taking those three efforts and wrapping them together and allowing the grid operator to have control over these resources so that these things stop being a nuisance and become an asset.

It’s a three-year project, kicking off right now with a fair amount of integration between companies. In the first year, a lot of the heavy lifting will be done from an integration perspective. This is really a software integration effort, getting different tech providers from big old-school infrastructure players to work with the new breed of technology providers that are operating more on the customer side and connecting those. You’ve got Siemens and Alstom running the energy management systems that the grid operators use here.

Then you’ve got Stem, a smart inverter company, cyber security, forecasting. A lot of that software integration will be done in the first year with actual operation of customer-sited resources being used by grid operators over the second year, and a lot of evaluation of the reporting and resulting analysis in the third year.

The cliché that Hawaii is a test bed for solar, storage and other distributed generation is a well-worn one. But while there is a lot of solar on the grid, each grid is an island, a close-loop. Are there limits to what we can take away from Hawaii’s experiences?  

TG: A number of folks are saying, ‘Let’s look at Hawaii because they’re facing a lot of the challenges today we’ll be facing in five years’ time’. I think that that’s true to some extent. You have grids that will be achieving high penetrations of renewables on the mainland, or outside the US. But I think that those large interconnected systems, they do have a different set of challenges to renewables integration, a different set of options.

There’s nowhere to go on an island grid. Each of the Hawaiian islands has its own requirements to stay in balance at all times and has its own energy management system with the grid operator sitting there on each island.

I think that there are some lessons to be carried over, particularly around some of the business model innovations that are going on, more so than around necessarily the immediate technical challenges, because you have such a different set of options on the transmission side on larger grids than you do here. Here, it’s like you have to balance the whole grid within the distribution system.

There are a number of efforts being played out by HECO but also the Hawaii public utilities’ commission (PUC, the state’s regulator), which is already taking a stance that customers need to be given an opportunity to participate in grid programmes and not just saying you can create your own kWh with a PV system. Customers who are investing in these new technologies – smart controllable inverters, fast acting energy storage, new technologies for demand response – those customers need to be given an opportunity to provide the grid the services that it needs to stay stable as we take off generators and mothball them. We need those services, the other things generators provide to keep the grid stable, not just the energy they produce but all the other ancillary services need to be backfilled and where are we going to get them? We’d rather not build a new bulk resource. Let’s figure out if we can provide those grid services from the distribution side and from the resources that customers are investing in, with their own money and take advantage of that so that the ratepayers benefit.

...there are some lessons to be carried over, particularly around some of the business model innovations that are going on, more so than around necessarily the immediate technical challenges, because you have such a different set of options on the transmission side on larger grids than you do here."

Tags: hawaii, stem, island grid, grid response, demand response, aggregation, sunshot initiative, renewables integration, distributed generation