The Negawatt: How Electricity That Never Was Becomes a Commodity

In January 2023, MEA, Bangkok's power distribution utility, began recruiting participants for a pilot program that pays customers to cut their electricity use for a few hours a day. The program was aimed at large commercial sites, and the terms ran as follows: for standing ready to shed at least 50 kW on the grid's command during the windows of 1:30-4:30 pm and 7:30-10:30 pm, participants received 44.57 baht per kilowatt per month, plus a separate payment for every kilowatt-hour actually saved. The mechanism being tested here is called demand response. Let's unpack why a state-owned utility would pay extra for electricity nobody consumed.

What demand response is

In 1989, the American energy analyst Amory Lovins coined the word negawatt – his name for a megawatt that never had to be generated, because someone cut their consumption at the right moment. Behind the wordplay lies the point that matters here: to the grid, a megawatt saved at the right moment is indistinguishable from a megawatt produced, which means it can be bought and sold like any other energy resource.

To see what this idea means in practice, remember that a power system is built not for average but for peak demand: for that evening hour when air conditioners, kitchens, and EV chargers are running in thousands of buildings at once. For the sake of a few hours a day, the system has to keep reserve power plants and reinforced networks that sit idle the rest of the time. This is expensive, and in the end every consumer pays for it through their electricity rates.

Demand response comes at the problem from the other end. Suppose the system is a hundred megawatts short at peak. There are two ways to get them: fire up a reserve plant, or ask consumers to shed those same hundred megawatts for a while. From the grid's point of view the result is the same – so let the power plants' megawatts and the consumers' negawatts compete at auction, and let the system buy whichever comes cheaper. A consumer who drops from 250 to 170 kW on the grid's command is selling it 80 kW, exactly as a power plant sells its output.

By now this is not just a thought experiment. Back in 2011, FERC, the U.S. federal energy regulator, ordered wholesale market operators to pay for verified demand reductions at the same price as for generated electricity. Power producers saw this as competition and took the case all the way to the Supreme Court, which sided with the regulator in 2016. Billions of dollars were at stake, and the outcome settled the negawatt's status as a marketable commodity.

What is happening in Thailand

Thailand is preparing to buy dispatchable load reduction on a permanent basis – and, by the look of it, fairly soon.

The pilot described above was sized at 50 megawatts. Dispatch was handled by EGAT, the national generation and transmission operator: in 2023 it opened the Demand Response Control Center, from which commands go out to participants. EGAT also acted as the buyer of the capacity, while the distribution utilities MEA and PEA served as load aggregators – Intermediaries who bundle demand reductions from many buildings into a single package and answer to the grid for delivering it. EGAT plans to go on to set up five regional centers and open the aggregator role to private companies.

Now the Energy Policy and Planning Office (EPPO) is working out a model for procuring demand response on a regular basis, as an alternative to building new power plants and a way to cut LNG imports. In August 2025 the National Energy Policy Council, the country's top energy policy body, reviewed the pilot's results, endorsed a roadmap for developing demand response, and instructed the Energy Regulatory Commission (ERC) to build future payments to participants into the base electricity tariff. In other words, the money to pay for load reduction will be collected from every consumer in the country – exactly the way everyone already pays to keep reserve power plants on standby. March 2026 also brought a dedicated body charged with developing demand response resources.

Load reduction targets are to be written into PDP2026, the new power development plan that runs to mid-century, alongside the power plants the plan lists by name. The country, in other words, intends to plan unconsumed megawatts in the same document, and in the same units, as built ones.

The rules are still being written: as of spring 2026, the study on which EPPO will base its specific procurement model was about halfway done. There are no actual rates or terms of participation for new customers yet – but it is a fairly safe bet that there will be.

What this might look like in the near future

Picture a large hundred-room hotel in Phuket, with solar panels, batteries, and EV chargers. On a normal evening it draws 250 kW from the grid. One day at six in the evening, a signal arrives from the aggregator: between 7:30 and 10:30 pm, the system needs the load brought down. The hotel's EMS (energy management system) takes stock: the batteries are at 82%, a 30% floor is enough for backup, and car charging can be throttled. The EMS starts discharging 50 kilowatts from the batteries, trims the EV chargers by 25 kW, and pushes another 10 kW of non-essential loads back to overnight. For those hours the hotel draws 165 kilowatts instead of 250, and the guests notice nothing.

To get paid, the reduction has to be measured and verified. The calculation runs automatically, from meter data. Payment is based on the difference between actual consumption and the baseline – an estimate of what the building would have consumed had there been no signal. The aggregator derives the baseline by formula from the building's consumption history on similar past days.

The equipment requirements follow from the logic of the scheme.

•      You need an interval meter that records readings every 15-30 minutes. (Thai customers in tariff categories 3, 4, and 5 have one anyway – it is required for the TOU tariff.)

•      You need an energy management system that takes the grid's reduction request and decides how to put together the promised kilowatts without touching guest comfort or the emergency reserve. It also keeps the log: how much the battery discharged, how much the panels produced, which loads were curtailed and when.

•      To cut its draw from the grid at the right moment, the building's own energy system needs real flexibility. That comes from the battery itself, and from sizable loads that can be shifted in time – EV chargers, for example.

•      You need a communication channel to carry the customer's data to the aggregator. In this scheme, data matter no less than the battery: a kilowatt that cannot be confirmed by measurement does not exist as far as the grid is concerned.

The important thing to understand is that the utility is paying first and foremost not for the energy saved but for guaranteed capacity during peak hours, when the alternative would be firing up a reserve plant. That is why the main money flow – 44.6 baht per kW per month, under the terms of the 2023 pilot – comes simply for standing ready to shed load on command. The second component (2.6 baht per kilowatt-hour in the pilot) pays for the energy actually saved during events. Our hotel would thus collect 3,800 baht a month, or around 45,000 baht a year, for readiness alone. How much the second component adds depends on how many events are called and how long they last.

As these numbers show, at the rate known today, demand response does not pay for a battery. Its main value still lies in raising solar self-consumption and sidestepping the peak tariff. Taking part in load reduction is a bonus on top of economics that already work – not a business in its own right.

Then again, the pilot rate was set administratively, for a one-off test with a closed pool of participants. Once payments become part of the base tariff, load reduction acquires annual targets, and private aggregators are let in, the price will be set under very different conditions.

For comparison: in America's ConnectedSolutions program, which operates in the region with the most expensive peak infrastructure in the country, a battery earns its owner about $200 per kilowatt over the summer season. Thai rates are nowhere near American ones, but the comparison shows that today's level of payment is not a ceiling.

Why it is worth thinking about now

An inverter and a battery last ten years and more, so equipment bought today will in all likelihood still be in service when the procurement mechanism goes live. Today, two systems of equal capacity may cost about the same – but with a demand response program up and running, they will behave differently.

One can generate energy, store the day's sun, discharge it in the evening hours, lower the bill, and keep critical loads running through an outage.

The other does all of that and is ready for demand response besides: its inverter and charging stations accept commands over open protocols, and its management system does more than optimize the owner's bill – It can take an aggregator's signal and report back the actual reduction.

The price difference at purchase is usually small – a matter of which inverter model and what class of EMS you choose. The difference in capability will show itself on the day a contract with an aggregator becomes genuinely available.

‍