Liberty Utilities will own up to 500 units, but the compromise allows third parties to participate, too.
Negotiations in New Hampshire produced a pivotal test case for how home energy devices can reduce overall grid costs.
Liberty Utilities proposed a groundbreaking home battery pilot last year, in which the company would own 1,000 Tesla Powerwalls in customer homes. The customers would get backup power and a time-based retail rate; the utility would aggregate the batteries to reduce its costs for monthly and annual system peaks and to offset wires upgrades.
That concept triggered a debate among New Hampshire stakeholders about the appropriate limits of utility ownership in a competitive marketplace. Companies like Sunrun wanted industry to have a chance to participate and get time-of-use rates for their customers. And questions remained about the utility’s ability to forecast peaks and respond to them.
A settlement has emerged from those negotiations, signed November 15. Regulators held a hearing on it Thursday and will finalize their decision in the coming weeks.
The consensus document scales down Liberty’s proposal and institutes granular checkpoints it must clear to scale up. It also creates a “bring your own device” program that other companies can compete for.
Sunrun didn’t sign on, but declined to oppose the outcome. It all turned out amicably, without the bloodletting that ensued from the net-metering battles of yore.
“It shows where we’re at in 2018: A small state [and] small utility can submit a proposal for residential batteries that really moves the entire industry forward,” said Chris Rauscher, Sunrun’s public policy director for the Northeast. “We’re no longer in a cost conversation, we’re in a value conversation, and that’s exactly where we should be.”
Assuming nothing derails regulatory approval, New Hampshire could become the proving ground for several of the hottest topics in distributed energy policy.
Pilot, but verify
The pilot taps into two grid edge trends: tackling the demand side of the equation to reduce peak expenses and leveraging private energy investments as aggregated grid resources, often referred to as virtual power plants.
In basic terms, customers will get subsidized batteries in exchange for letting the utility use them to fight system peaks. The settlement narrows the scope of the original proposal, but retains the same aspirations.
Liberty wanted to own 1,000 Powerwalls. The settlement allows an initial tranche of 200, and the utility can unlock another 300 if the first group performs appropriately.
To host the batteries, customers pay either $2,433 per system upfront, or $25 per month for 10 years (with a free extension at the end for another five years). That’s a bit pricier than Vermont utility Green Mountain Power’s $15-per-month Powerwall offering, but well below the sticker price for the gadget, which became considerably more expensive this year.
Customers get to use their batteries as they see fit, except when Liberty predicts a peak event and gives 24 hours’ notice. Then the company takes control, charges the battery on cheap late-night power, and discharges it as needed during the peak window to reduce the total draw from ISO New England (and the customer’s bill during the most expensive time of day).
Whereas Green Mountain Power gives customers the benefit of backup power, the time-of-use rate, with its critical peak rate and low nighttime charge, incentivizes shifting the home’s energy usage in ways to benefit the grid at large.
“Customers can use the battery for time-of-use arbitrage to offset their upfront investment,” said consultant Lon Huber, who helped the ratepayer advocate design the rate. “We looked statistically at when the peaks are likely to occur, and aligned the TOU peak period around system peak and transmission peak.”
That’s all the more notable given the dearth of time-based rates in New England, which frustrates efforts to enlist customer behavior in the reduction of system peak costs.
To unlock Phase 2, Liberty has to prove that the actual battery performance matched expected peak hour demand reductions with at least 75 percent accuracy. Real-world cost savings from the program also have to match the projections in the cost-benefit analysis.
The message to Liberty: Show that this works and we’ll let you buy more things.
Hitting a peak is not a negligible task. The utility must harness whatever data it has to determine when it might occur. Then it must allocate stored energy capacity while accounting for the fact that it may face even greater demand in the next hour.
By taking control of the batteries, Liberty ensures the customer doesn’t drain the charge before a peak hour arrives. Still, the utility must dispatch wisely, so it doesn’t burn its capacity in one hour only to discover the next will be an even bigger peak.
This peak-shaving differs from the kind practiced by commercial storage operators like Stem and AMS. Those companies serve their customers first and foremost, and seek to improve returns by bidding into the markets or serving utility contracts. Those dual goals at times work at cross purposes, hence the need for a statistically savvy fleetwide approach.
The contents of Liberty’s Phase 1 evaluation will provide an unprecedented peek into the actual operational effectiveness of distributed batteries, which normally resides in obscurity behind the veil of “trade secrets.”
Join the party
Sunrun didn’t succeed in stopping utility ownership of the batteries outright, but the settlement added a pathway for third-party aggregators to participate.
The deal would form a working group to figure out the details, but the gist is that companies besides the utility would be free to pitch customers on getting a battery from them and participating in peak events on behalf of all ratepayers. Who can participate and how they will be compensated are yet to be determined.
This option expands the potential scope of the program.
Instead of using ratepayer money to socialize the cost of the batteries, which in turn reduces system-wide costs, the BYOD structure leverages private investment for public benefit. If the private aggregators reduce peak charges, they get a cut of the savings; if they don’t perform up to snuff, they didn’t waste the public’s money in the process.
Just like Liberty will subsidize the cost of the Powerwalls, expected revenue from grid services lets companies like Sunrun lower their upfront pricing.
“The more we can do value stacking, the lower the price goes for storage, and that means increased access down the income spectrum,” Rauscher said.
There’s no reason this model would have to end with storage. It could well expand to smart thermostats, controllable heating and cooling, and other smart energy devices, creating a robust case study in support of demand-side management.
Ultimately, the more people who access advanced price signals, the better, Huber said.
“If a third-party aggregator can handle the top hour of the month, top hour of the year, by all means they should grab it, as long as there’s no cost shift,” he noted.
The settlement posits one way to resolve the tension between sophisticated price signaling and customer confusion. Accurately pricing the cost of electricity production will make the grid more cost-efficient compared to the status quo, but it poses a real challenge to all those customers used to flat rates, who lack the tools or understanding to manage time-based charges.
The New Hampshire arrangement introduces more sophisticated price signals to residents via the time-of-use rate, but shields them from harm by letting the utility handle peak prediction and response. The third-party aggregators, too, take on the task of predictive analytics in order to deliver bill savings or backup power to their customers.
If this program makes customers happy and saves money for the system overall, its name will ring out at many a conference panel as a vision of the possibilities of distributed energy. Getting to that point just requires regulatory approval, actual deployment and rigorous execution on the concept.