Think B.I.G. Hawaiʻi← All toolsEvery year Hawaiʻi's grids turn away clean electricity that was already built, already paid for, and ready to flow. Smart Imports requires the utility to offer that surplus to flexible loads at its net cost, which is often zero. The payoff runs both ways: far more local green energy, because solar and wind that would have been curtailed gets used and new projects pencil without the curtailment penalty, and far cheaper power for utility customers, because every stored kilowatt-hour of free midday sun replaces an oil-fired one at the evening peak, and that fuel saving flows through to every bill.
This page is a working model for legislators and staff, not a report. It shows how much clean energy is being turned away today, lets you change every assumption, and shows what Smart Imports would be worth to drivers, ratepayers and the grid. The scene above is the goal: midday sun, panels at full power, nothing wasted, and no oil burned tonight.
Midday surplusUtility plants ordered to back off, and rooftop systems throttled once the home battery is full. Built, paid for, and wasted.
Result tonightLess oil burned at the evening peak, which lowers the fuel charge on every customer's bill; near-free daytime energy for whoever shifted; and more local energy stored on-island.
Hawaiian Electric reports "curtailment" narrowly: megawatt-hours it ordered utility-scale wind and solar plants to stop producing. By that measure, curtailment peaked at 88 GWh in 2021 and fell to 28 GWh in 2025 after large batteries came online on Oʻahu and Maui2. On Maui most of it is plain oversupply: midday generation exceeding what the island can absorb3.
That is only the visible part. A second, larger pool is energy that rooftop systems are not allowed to send to the grid at all. At the end of 2024, Hawaiian Electric had 246 MW of customer solar enrolled in programs that forbid daytime export or pay nothing for it (Customer Self-Supply, Standard Interconnection Agreement, Smart Export, and the new Non-Export program)4. Once the home battery is full, those panels are throttled. Nobody reports that energy as curtailed, but it is.
A third pool is coming. The PUC approved a procurement of roughly 1,650 GWh/yr of new variable renewables for Oʻahu and Hawaiʻi Island, in service 2031 to 20347. University of Hawaiʻi modeling shows that at those penetration levels, curtailment climbs from a few percent to around 20% of solar output unless the grid finds more daytime load8.
This page defines curtailment the way a ratepayer would: any clean energy that could be delivered to the grid but is not, because of the utility's dispatch choices, stability limits, or restrictions on distributed generation.
GWh per year, 2015–2025. Source: Hawaiian Electric key performance metrics2. Hover for detail.
| 2025 | Curtailed GWh | Output GWh | Share |
|---|---|---|---|
| Oʻahu | 14.3 | 861 | 1.6% |
| Maui | 10.7 | 301 | 3.4% |
| Hawaiʻi Island | 3.1 | 300 | 1.0% |
| Hawaiian Electric total | 28.1 | 1,462 | 1.9% |
Each grid throws away clean energy in its own way, and each has its own case for cheaper bills. On every island the same pattern holds: what HECO reports as curtailment is the smallest part of the surplus.
GWh. Reported is Hawaiian Electric's curtailment total2. Broad is what this page counts: curtailment the grid ordered, plus rooftop solar that is barred from exporting and gets throttled. The rooftop estimate follows the model's sliders below4. Owner-requested cuts are left out because they are not the utility's choice. Solar-plus-storage dispatch-down is not published, so it is marked, not estimated.
Reported curtailment on Oʻahu was 14.3 GWh in 2025, 1.6% of curtailable output, and 5.5 GWh in the first half of 20262. Nearly all of it, 13.6 GWh, is classed as "system constraint": local grid limits at the older wind and solar farms in HECO's logs, such as Kahuku, Nā Pua Makani, Kawailoa, Waihonu, Kalaeloa and West Loch3,24. Island-wide oversupply peaked at 22 GWh in 2021 and was down to 0.6 GWh in 2025, after the Kapolei battery came online3,16.
That is the narrow measure. Oʻahu's newest solar-plus-storage plants, Hoʻohana (52 MW / 208 MWh), Waiawa Phase 2 (30 MW / 240 MWh) and Mountain View (7 MW / 35 MWh), all added in 20255, do not appear in the curtailment logs24, so energy they are dispatched down goes unreported. The largest pool is on rooftops. At the end of 2024, 177 MW of Oʻahu customer solar was in programs that bar daytime export or pay nothing for it, including 476 Standard Interconnection systems averaging about 280 kW: warehouse and commercial roofs sized to their own load4. At the model's defaults that is about 59 GWh a year of throttled clean energy, four times the reported curtailment.
Renewables and rates. Oʻahu is furthest behind, at 32.3% RPS in 2025 against a 40% target for 20305. It also has most of the state's load and EVs, so it is where Smart Imports could move the most energy. Every surplus kilowatt-hour used at noon or stored for the evening offsets oil, and oil is what drives the 30.3¢ energy cost recovery charge on every Oʻahu bill. HECO credits a noon export at 13.5¢ while selling power at 47.8¢. Smart Imports would let customers buy midday surplus at close to what the utility itself says it is worth.
2025, GWh. Rooftop estimate follows the model sliders.
| MWh unless noted | 2024 | 2025 | H1 2026 |
|---|---|---|---|
| Reported curtailment | 25,204 | 14,282 | 5,485 |
| oversupply | 6,226 | 561 | 625 |
| system constraint | 18,319 | 13,560 | 4,722 |
| facility requested | 659 | 161 | 131 |
| Hybrid plants dispatched down | not reported | ||
| Export-barred rooftop solar | 176.9 MW, about 59 GWh/yr (est.) | ||
| Rooftop solar vs hosting capacity | 746 of 1,000 MW used (Q2 2026) | ||
Sources: curtailment workbooks2,3; rooftop program MW4; hosting capacity17. The 59 GWh figure is ours, at the model's 19% capacity factor and 20% throttled share. The category workbook's H1 2026 split sums to 5,478 MWh.
Maui has the highest curtailment rate of any large Hawaiian Electric grid: 3.4% of curtailable output in 2025. It is almost pure oversupply, 94% of the total in 2025 and 96% in 2024, meaning more wind and sun than the island can use3. Reported curtailment fell from 36.4 GWh in 2024 to 10.7 GWh in 2025, the first full year of AES's Kūihelani plant (60 MW / 240 MWh, 8¢ per kWh)21. It is climbing again: 7.6 GWh in the first half of 2026, already 71% of all of 20252.
Much of it is wind at night. HECO's logs show Kaheawa Wind II curtailed from about midnight to 6 a.m., and HECO notes that curtailment signals sent to some facilities at night are not recorded at all24. Kūihelani never appears in the logs. Maui's rooftop grid is nearly full, with 161 MW executed against 172 MW of hosting capacity17, and Waena, the island's first standalone grid battery (40 MW / 160 MWh), is not due in service until 202722.
Renewables and rates. HECO values a Maui noon export at 6.6¢, the lowest in the state, while residential customers pay 46.8¢23,6. That 40-cent gap is the clearest affordability case in Hawaiʻi for at-cost surplus sales, by day and by night. Hosting capacity is set largely by how little load there is at midday, so every EV, water heater and farm pump that moves to noon also makes room for more rooftop solar on the same wires.
2025, GWh. Rooftop estimate follows the model sliders.
| MWh unless noted | 2024 | 2025 | H1 2026 |
|---|---|---|---|
| Reported curtailment | 36,447 | 10,732 | 7,585 |
| oversupply | 35,162 | 10,131 | 6,874 |
| system constraint | 1,215 | 566 | 390 |
| facility requested | 70 | 35 | 322 |
| Hybrid plants dispatched down | not reported | ||
| Export-barred rooftop solar | 32.8 MW, about 11 GWh/yr (est.) | ||
| Rooftop solar vs hosting capacity | 161 of 172 MW used (Q2 2026) | ||
Sources: curtailment workbooks2,3; rooftop program MW4; hosting capacity17. The 11 GWh figure is ours, at the model's defaults. The category workbook's H1 2026 split sums to 7,586 MWh.
The Big Island's reported curtailment is the smallest of the three big grids: 3.1 GWh in 2025, 1.0% of curtailable output2. Seven-tenths of it was "facility requested", mostly from Pakini Nui, the kind of cut a wind farm makes to protect its turbines in high winds. That is not the utility's choice. Only about 0.9 GWh was ordered by the grid operator for oversupply or system constraints3.
The event logs show where the pressure is. Pakini Nui, Hāwī and small solar sites are cut back between about 10 a.m. and 3 p.m. on sunny days, and Pakini Nui is also curtailed for oversupply between 2 and 5 a.m., when load is lowest and the generators the utility keeps running for stability already fill the grid19. Surplus is not only a daytime problem.
Waikoloa Solar and Hale Kuawehi never appear in those logs. Both are 30 MW solar plants with 120 MWh batteries under dispatchable contracts, so when their batteries are full and the grid has no room, Hawaiian Electric simply dispatches them down. That energy is not reported as curtailment anywhere14,20. The 2025 RPS report suggests the loss is modest so far: utility solar on the island delivered 131 GWh, roughly a 27% capacity factor on the two plants' 60 MW once Hale Kuawehi's March start is allowed for5. The pressure is building, though. Hawaiian Electric's own Q2 2026 report shows 157.5 MW of rooftop solar on the island against a hosting capacity of 119.1 MW, with the Hāmākua thermal plant assumed running17.
Renewables and rates. The Big Island leads the state at 57.3% RPS5, but its residential customers still pay 43.1¢ while HECO credits a noon export at 10.6¢6,23. Here the lever is the rooftop fleet that is already over the grid's hosting limit: daytime load that soaks up its surplus lets more of it run, and overnight wind sold at cost would cut what EV owners pay to charge at home.
2025, GWh. Rooftop estimate follows the model sliders.
| MWh unless noted | 2024 | 2025 | H1 2026 |
|---|---|---|---|
| Reported curtailment | 1,918 | 3,084 | 776 |
| oversupply | 954 | 367 | 70 |
| system constraint | 120 | 540 | 31 |
| facility requested | 844 | 2,176 | 675 |
| Hybrid plants dispatched down | not reported | ||
| Export-barred rooftop solar | 37.0 MW, about 12 GWh/yr (est.) | ||
| Rooftop solar vs hosting capacity | 158 of 119 MW: over the limit (Q2 2026) | ||
Sources: curtailment workbooks2,3; delivered energy from the 2025 RPS report5: wind 145,383 MWh and utility solar 131,027 MWh in 2025. The 27% capacity factor and 12 GWh figures are ours.
Defaults come from the sources at the bottom of the page. Every slider is yours to move; the results update as you go. Figures are annual, Hawaiian Electric islands only (Kauaʻi's co-op is not included).
| Use | GWh | Savings / yr | Versus |
|---|
Surplus = reported utility-scale curtailment + (export-barred rooftop MW × 8,760 h × capacity factor × throttled share) + (new renewables × curtailed share). EVs: each kWh drives 3.3 miles (0.30 kWh/mi) and displaces gasoline at 25 mpg11; savings = gasoline avoided − import price paid. Batteries: kWh charged × round-trip efficiency displaces an oil-fired kWh that evening at the fuel cost you set; this saving flows to every ratepayer through the fuel clause. Thermal, ag and other loads: treated as load shifted from other hours, so each kWh saves (retail − import price) on the bill and displaces an oil-fired kWh. Fuel oil: 0.08 gal per kWh (EIA)12. CO₂: 2.46 lb per oil-fired kWh (EIA)12 and 8.89 kg per gallon of gasoline (EPA)13. "Days of fossil generation" uses HECO's 2025 fossil output of 6,740 GWh (10,668 GWh generated minus 3,928 GWh renewable)5.
Four situations that are stuck today and unstuck once the grid is obliged to take surplus and pass it on at cost. All four sit on the bill's priority list: essential services, households, farms and public charging.
A fire station installs a solar array about three times the size its daytime load needs, plus a battery, on a microgrid that can island. On an ordinary day two-thirds of the output is surplus. In a storm week, when the grid is down and the sky is grey, a system that size still covers the whole station: bay doors, radios, refrigeration, lights, the works.
A homeowner wants a solar system big enough to run her whole house. Under today's non-export rules the only way to make the midday surplus useful is to buy more onsite storage, so the second battery, not the panels, becomes the cost that stops the project. In the first half of 2026 Hawaiian Electric customers installed 55 MW of batteries against 36 MW of solar18.
A Puna or Hāmākua food hub runs cold rooms, a flash freezer, an ice machine for the fishing fleet and irrigation pumps. All of it is flexible: a cold room can be pre-chilled a few degrees at noon and coast through the evening; tanks can be filled at midday; ice can be made whenever power is cheapest.
School buses sit parked from 9 a.m. to 2 p.m., exactly when the surplus peaks. So do county trucks, rental fleets, and the cars of everyone at work. A depot or workplace charger that fills vehicles at cost in those hours turns the state's 42,299 EVs10, and every bus that follows, into the grid's largest flexible load.
SB 3163, introduced by Sen. Gabbard with Sens. San Buenaventura, Lee and McKelvey in January 2026, was referred to CPN/EIG and WAM and did not get a hearing1. The 2027 version keeps the same architecture:
Two changes we recommend for 2027, both drawn from the Big Island data above:
Hawaiian Electric's modern solar-plus-storage contracts pay developers a monthly lump sum based on the plant's energy potential and availability, not on energy actually delivered14. Ratepayers have already paid for every curtailed megawatt-hour. Selling that energy at the utility's net cost returns value that would otherwise be thrown away; it takes nothing from other customers. And the savings are not only for participants: fuel is the largest single line on a Hawaiʻi bill, 22 to 30 cents of every kilowatt-hour in May 20266. Every battery, water heater or farm that fills up on free midday sun instead of running at the oil-fired evening peak shrinks that fuel charge for every customer on the island.
The PUC found that the 565 MWh Kapolei battery, Oʻahu's largest, is "mostly powered by fossil fuel generation plants" and asked the utility to plan to minimize that15. Smart Imports gives batteries, water heaters and EVs a price signal to fill up on midday sun instead. Every kilowatt-hour stored from local sunshine is one that does not depend on a tanker arriving on schedule.
Utility-reported curtailment was 28 GWh in 2025, down from 88 GWh in 2021, because Kapolei Energy Storage and Maui's new batteries absorb midday surplus2,16. But that number counts only utility-scale plants. It excludes the 246 MW of rooftop systems forbidden from exporting at midday, and it says nothing about the 1,650 GWh of new variable generation the PUC just approved. Set the first slider to 28 and the rest of the model still shows a large resource.
The utility earns no revenue on energy it curtails or refuses to accept. Smart Imports monetizes energy at the utility's net cost, so fixed costs are unchanged, and the new metered load pays the same customer and delivery charges as any other service. The bill requires the PUC to design the cost formula and to prevent gaming in either direction.
We should, and the model treats midday battery charging as one of the biggest uses. But batteries alone do not fix it: UH modeling found Oʻahu batteries cycle less than once a day on average, and curtailment persists because evening and overnight load is limited8. EVs, water heaters and farms are load that can move to midday at almost no capital cost.
No. The bill directs the PUC to make Smart Imports complement those programs. Today's Shift-and-Save TOU rates and the Smart Export daytime credit already point in this direction; Smart Imports makes the daytime price reflect the utility's true marginal cost, which is near zero when plants are being curtailed.
The bill requires a transparent cost formula, independent verification, public disclosure and quarterly reporting of surplus periods, prices, enrollment and curtailment impact, with PUC penalties for misreporting1.