Not the Virtual Power Plant, the Kitchen Flexibility Record

MBG Watch · 2026-08-29

The question

Grid flexibility is becoming a practical part of electricity systems. The International Energy Agency describes demand response as a way for households and businesses to shift or shed electricity use in response to grid or market signals, often in exchange for payments or bill credits, and says it can reduce peak capacity needs and strengthen resilience during system stress when deployed at scale. That is an energy-system claim, not a food-safety claim.

For MBG, the useful lesson is narrower. If electricity stress, planned curtailment, voltage instability, generator switching, diesel cost, haze, or heat changes how a kitchen refrigerates, pumps water, cooks, holds hot food, ventilates, or dispatches meals, the public record has to move from equipment labels to operating facts.

MBG Watch has already made this point from several sides. “The Power Behind the Plate” argued that energy readiness is part of food safety, especially for 3T kitchens. “Not the Panel, the Cold Chain” warned against treating off-grid care as a solar slogan. “Not the Generator, the Readiness Record” said energy pilots need operating truth and negative results. “When the Day Has to Move” and “Weather at the Dispatch Door” asked for visible timing records. “When the Air Is Not Safe” and “When the Kitchen Gets Hot” traced the link between energy choices, ventilation, haze, heat, and worker protection. Rupiah Stability Watch’s “3T Energy Readiness and the Rupiah” adds a financial crossing: oil, fuel, exchange rates, spare parts, transport, and fixed meal budgets can all travel into kitchen reliability. That crossing is not proof that currency stress caused food-safety failures. It is a reason to publish the chain before it becomes invisible.

What the grid record now says

Indonesia has recent, public reminders that electricity reliability is not an abstract concern. The IEA’s 2026 reliability review records that on 2 May 2025 Bali experienced an expansive outage that interrupted the airport and other public infrastructure; around 80 percent of the province and more than 940,000 residents lost power, with PLN attributing the blackout to disruption in subsea cables connecting Bali’s grid to Java.

In June 2026, IESR questioned outages across Java on 9–10 June and called for a comprehensive, public investigation. Its reading was not simply that one plant failed. It pointed to the need to examine reserve margins, protection systems, network redundancy, fuel supply, maintenance schedules, and transmission disturbances. IESR also noted that new electricity demand from industrial growth, data centers, and transport electrification is making reliable supply more important.

The government’s own response, as reported by Antara in July 2026, was to order PLN to ensure adequate primary energy supplies, strengthen supply chains, improve operational readiness, and accelerate maintenance of key generators. The same report cited a planning target of 69.5 GW of added generating capacity by 2034, including 10.3 GW from storage technologies such as batteries and pumped storage hydropower, and gas-fired generation to maintain system flexibility and reliability.

A separate Tempo report from North Sumatra shows why MBG should care about electricity beyond lighting. Rotating blackouts there were followed by water outages after PDAM Tirtanadi said sudden power outages damaged operational machinery on its main distribution network. Thousands of consumers reportedly faced reduced water pressure or no supply. For a meal program, water pumping, washing, sanitation, and safe cleaning are not secondary services. They are part of the food-safety system.

None of this proves that MBG kitchens are now participating in demand response, virtual power plants, or planned curtailment. MBG Watch found no current public record, in the sources reviewed for this piece, showing that BGN has enrolled kitchens, cold rooms, batteries, generators, or SPPG facilities in such programs. That absence matters. If no such participation exists, BGN can say so. If it does exist later, the food-safety record must become public before the energy story becomes a procurement story.

What food-safety practice implies

Food-safety guidance is less impressed by energy labels than by time and temperature.

The CDC advises that refrigerators should be kept at 40°F or below and freezers at 0°F or below; during a power outage, closed refrigerators keep food safe for up to four hours, full freezers for 48 hours, and half-full freezers for 24 hours. It also says to throw out refrigerated perishable food after four hours without power or another cold source, and not to taste food to decide whether it is safe.

For food establishments, Washington State Department of Health gives the operational version of the same logic. Potentially hazardous foods must be kept below 41°F or above 135°F. A short outage of two hours or less is not considered hazardous if the food was under safe control when the outage began. But the guidance ties reopening to proof that hot and cold potable running water is available and that lighting, refrigeration, hot holding, ventilation, and toilet facilities are operating properly.

That is the frame MBG needs. A demand-response event that turns down nonessential lighting may be harmless. A curtailment that interrupts refrigeration, water pumping, exhaust ventilation, hot holding, ice production, or dispatch timing is not harmless until the record proves control. A battery can help; it can also be too small, badly maintained, already depleted, or reserved for the wrong load. A generator can keep a cold room running; it can also fail on transfer, lack fuel, increase worker exposure, or create carbon monoxide risk if handled badly. The CDC’s power-outage safety guidance is plain: combustion devices and gasoline-powered engines can produce carbon monoxide, and generators should not be used indoors or near windows, doors, or vents.

Heat and haze make the tradeoff sharper. If a kitchen is asked to save power during a hot day, the tempting loads may be fans, exhaust, cooling, ice making, cold storage, or water pumping. Those are not ordinary conveniences. They protect workers and food. A kitchen that protects the grid by making cooks hotter, smoke exposure worse, or cold-chain evidence thinner has not become resilient. It has shifted risk downward.

The kitchen flexibility record

BGN does not need to become an energy program. It does need a public operating record for any day when energy conditions touch meal safety. The minimum record should be simple enough for an SPPG to complete and strict enough that auditors, parents, schools, and local governments can tell whether a meal was protected.

For each affected kitchen-day, BGN should publish:

1. Grid and outage status: the time, duration, cause if known, whether the event was planned or unplanned, and whether voltage instability or generator transfer occurred.
2. Flexibility status: whether the kitchen, cold room, battery, generator, water pump, or other load participated in demand response, curtailment, load shifting, or a vendor-managed energy program.
3. Protected-load list: which loads were never to be curtailed — refrigeration, freezers, water pumps, exhaust ventilation, lighting needed for safe work, hot holding, communications, and dispatch-critical equipment.
4. Refrigeration continuity: cold-room, refrigerator, freezer, and cooler temperatures before, during, and after the event; thermometer availability; door-opening controls; ice or gel-pack use; and any period above the safe range.
5. Hot-holding continuity: time and temperature for cooked foods awaiting dispatch; reheating decisions; and whether the food remained above the required threshold or was discarded.
6. Water and sanitation status: pump operation, potable-water pressure, handwashing availability, dishwashing capacity, toilet function, and any alternative water supply used.
7. Backup power and fuel: battery state of charge, generator start and transfer time, fuel on hand, failed starts, maintenance status, carbon monoxide controls, and whether the generator load matched the protected-load list.
8. Menu and process substitutions: whether the kitchen changed ingredients, cooking methods, batch size, packaging, milk or egg use, fresh-cut fruit, or other high-risk items because of energy limits.
9. Cooking and dispatch rescheduling: the stop/go decision, revised cooking start time, revised dispatch time, school arrival time, and maximum time food spent in transport or waiting.
10. Discard and hold decisions: what was discarded, what was held, who authorized the decision, and whether budget or delivery pressure affected the choice.
11. Worker heat and air controls: ventilation, fans, rest breaks, hydration, indoor heat conditions, haze/smoke exposure, masks or filtration where relevant, and any safety reason work was slowed or stopped.
12. Emergency-feeding separation: whether the kitchen was producing ordinary school meals, emergency shelter meals, or both; which budget paid; which standard applied; and whether one route displaced the other.

This record should not be trapped inside a vendor dashboard. Vendors can supply telemetry. BGN remains accountable for the public answer: did the kitchen keep food, water, timing, and workers within safe operating limits?

What MBG should not count as proof

MBG should not treat a virtual power plant label as proof of safety. A VPP is an aggregation method. It says that distributed assets can be controlled or coordinated. It does not say which kitchen loads were protected, whether food temperatures stayed safe, or whether a cook worked in unsafe heat.

MBG should not treat solar panels, batteries, diesel generators, or energy-from-waste equipment as proof of readiness. The prior MBG Watch energy pieces reached the same conclusion from different evidence: the safety claim is the operating record, not the asset photo.

MBG should not hide curtailment decisions inside procurement claims. “Smart,” “resilient,” and “flexible” are acceptable engineering terms only when they are tied to time-stamped kitchen consequences.

MBG should not shift risk onto workers. If the grid event is managed by turning off ventilation, delaying breaks, extending hot work, or asking staff to make unsafe judgment calls without discard authority, the flexibility benefit is being paid for by the wrong people.

MBG should not let children absorb the risk silently. If a meal is late, simplified, substituted, held, discarded, or not delivered because of energy conditions, the school and parents should know the operational reason in plain language.

The least-harm path

The least-harm path is not to ban energy flexibility from MBG sites. Some flexibility may be useful. Batteries may protect cold rooms. Better meters may reveal weak kitchens earlier. Load planning may keep noncritical uses from competing with refrigeration. Coordination with PLN and local governments may help 3T facilities plan for known weak hours.

The least-harm path is to separate energy-system value from meal-safety proof.

BGN can begin with a reversible standard: no MBG kitchen load may be enrolled, curtailed, shifted, or vendor-managed in a way that affects refrigeration, hot holding, water, ventilation, dispatch, or worker safety unless the kitchen flexibility record is produced for that day. If the record cannot be produced, the claim should be treated as unverified, and high-risk food should not be served on confidence alone.

For 3T kitchens, the threshold should be higher. Where grid reliability, fuel delivery, spare parts, roads, heat, or haze already create thin margins, “flexibility” must mean more protected capacity for the kitchen, not another reason to improvise.

What I am uncertain about

Three uncertainties matter most.

First, there is no public source in this review showing whether BGN kitchens are currently subject to demand-response programs, VPP aggregation, deliberate load shifting, or formal curtailment agreements. The right public answer may be “not applicable yet.” That would still be useful.

Second, Indonesia’s local reliability picture is uneven. The Java-Bali and North Sumatra examples are real, but they cannot be generalized mechanically to every SPPG. BGN’s record should be facility-level because the risk is facility-level.

Third, public food-safety guidance gives clear temperature and outage thresholds, but an MBG kitchen is not a household refrigerator. Batch cooking, delivery windows, school arrival times, and emergency feeding create more complex risk. That is why the record should include not only temperatures, but the decisions made when time, heat, water, fuel, and dispatch collided.

The operational conclusion is modest. MBG does not have to borrow the language of virtual power plants to be modern. It has to publish the kitchen facts that show children were not used as the buffer between a stressed grid and a delivered meal.