MBG Kitchens in the Rupiah Energy Ledger: Flexible Load, Essential Load, or Both

Rupiah Stability Watch · 2026-08-30

The premise

Indonesia is beginning to assemble the instruments that make electricity flexibility more measurable: rooftop solar quotas, advanced meters, digital applications, and a public discussion about grid reliability. At the same time, the Makan Bergizi Gratis network is becoming a large, daily operating system: kitchens, water, cold storage, cooking fuel, workers, vehicles, dispatch schedules, and food-safety decisions.

Those two ledgers now touch. A kitchen can sometimes move electricity demand away from a stressed hour. It can also be an essential load that should not be interrupted. The rupiah question is not whether MBG kitchens can become a virtual power plant. That would be too broad. The narrower question is how to classify kitchen loads before a grid-stress event, fuel disruption, or curtailment order turns a technical choice into a food-safety and fiscal cost.

Rupiah Stability Watch has treated this as an operating-ledger problem in earlier work: “Virtual Power Plants and the Rupiah” examined household flexibility and fuel-import exposure; “3T Energy Readiness and the Rupiah” and “Power Reliability, Sleep, and the Rupiah” examined reliability as a productivity and wellbeing channel; “MBG Operating Disclosure and the Rupiah” and “Hidden Inflation in the Meal Tray” examined how meal-program costs can become invisible when they are not recorded. MBG Watch’s “Not the Virtual Power Plant, the Kitchen Flexibility Record,” “The Power Behind the Plate,” and “Not the Panel, the Cold Chain” point to the same boundary from the food-safety side.

The public record supports a careful reading: MBG kitchens are both flexible and essential, but at different layers of the operation.

What flexibility can safely mean for a kitchen

Flexibility is safest when it is pre-classified, bounded, and visible. For an SPPG kitchen, the flexible category is not “food service” as a whole. It is a set of non-safety tasks that can be moved in time without weakening the meal.

Examples include dishwashing outside the peak interval where clean inventory is adequate; non-critical water pumping where safe storage exists; battery charging; ice-making; some non-critical preparation; refrigeration pre-cooling before a predictable peak; and deferrable equipment cleaning that does not collide with hygiene standards or the next production cycle.

Indonesia’s rooftop-solar regulation shows why this classification matters. The Cabinet Secretariat’s summary of MEMR Regulation No. 2/2024 says rooftop systems are now managed through PLN-linked quotas and application-based procedures, and that excess exported electricity is no longer counted in the customer bill. It also states that advanced meters are used as substitutes for import-export kWh meters and that rooftop capacity must account for system reliability. ABNR’s legal summary reads the same regulation as replacing the old capacity cap with a quota system, abolishing net metering, and requiring “advanced meters” with two-way communication and measuring capabilities.

That is not yet a full demand-response regime for kitchens. It is a measurement and interconnection base. It can tell policymakers where flexible loads might later be recorded. It does not by itself say which food-service loads can safely respond to grid stress.

The useful first step is therefore not curtailment. It is a kitchen flexibility record: what load was shifted, when, by whose authority, under what food-safety condition, and with what effect on LPG use, generator fuel, water status, cold storage, and dispatch time.

What should be protected from curtailment

The protected category is larger than the refrigerator alone.

BGN’s 2026 technical guideline, retrieved from the official BGN document repository, describes SPPG as the operating unit for free nutritious meals and sets an intended national scale of 35,000–40,000 SPPG, including 8,000–10,000 in remote areas. The same guideline says a site should be within a maximum 6 km radius or 30 minutes of target beneficiaries, have PLN electricity access, and have clean logistics access. It requires electricity installation to meet safety standards, electricity payments to be orderly, and backup power from a generator. It also requires gas infrastructure, permits LPG, LNG, natural gas, or biogas, and specifies that LPG gas stations should be separated from food-processing rooms, with 50 kg LPG cylinders preferred where available and 12 kg cylinders allowed where 50 kg cylinders are unavailable.

Those details are operationally important. They mean a kitchen is already a coupled energy system: grid power, generator backup, gas cooking, water, sanitation, communications, CCTV, logistics, and timed dispatch.

Critical loads and functions should therefore include refrigeration and temperature control; safe water and pumping where water is not already stored safely; cooking completion once a batch has entered a safety-sensitive stage; hot holding and cooling steps; ventilation where heat, smoke, or gas safety are present; dispatch timing; emergency feeding windows; and communications or monitoring systems needed to document safe operation.

International food-safety guidance supports this caution. WHO’s 2024 publication on power outages states that emergencies and natural disasters frequently cause outages that can compromise food quality and safety, and that guidance is needed not only for households but also for food suppliers, vendors, inspectors, and workers. The public U.S. food-safety search records reviewed for this piece give a concrete benchmark: refrigerated food is generally considered safe for about four hours without power if the refrigerator remains closed; after that, perishable food may need to be discarded. That benchmark is not an Indonesian regulation and should not be imported mechanically. But it illustrates the principle: food safety is time- and temperature-bound, so a load-control rule that ignores temperature control is not flexibility. It is hidden spoilage risk.

The rupiah transmission chain

Rupiah weakness enters this ledger through four channels.

First, fuel imports. Indonesia still has material LPG import dependence. Antara reported from the June 2026 Tuban Mini LNG Plant inauguration that Indonesia consumes more than 8.5 million metric tons of LPG annually, produces about 1.91 million metric tons domestically, and imports the remaining 7.47 million metric tons. That is the external-balance link. Every kitchen that uses LPG sits partly inside an imported-fuel ledger, even when the meal budget is denominated in rupiah.

Second, fuel substitution and scarcity. Antara reported in August 2026 that the Energy and Mineral Resources Ministry was testing 3 kg compressed natural gas cylinders as an alternative to subsidized 3 kg LPG. The same report said CNG raw material is available from domestic natural gas, while LPG import dependence remains a policy concern. That does not prove CNG is ready for SPPG kitchens; the report itself notes technical pressure differences between CNG and LPG. It does show why kitchen energy records matter. A rupiah-stressed subsidy system will look for substitutes, but substitutions that are not operationally safe or available can move costs into service failure.

Third, equipment and spare parts. Cold storage, generators, controls, meters, modular kitchen components, and vehicles can carry imported components or foreign-currency-linked replacement costs. Liputan6, citing the Public Works Ministry’s SPPG construction push, reported an estimated Rp4.5 billion cost for one SPPG kitchen including the main kitchen, parking, clean-water network, wastewater treatment, kitchen equipment, and food-distribution vehicles. Tradecorp’s discussion of SPPG kitchen layouts, while commercial rather than regulatory, usefully names the same physical complexity: 300–400 square meter kitchen options, production and packing rooms, ventilation, wastewater treatment, anti-slip surfaces, and modular construction.

Fourth, waste and compensation. If grid stress leads to generator running, the cost becomes diesel or gasoline. If LPG scarcity delays cooking, the cost becomes staff time and route disruption. If cold-chain control fails, the cost becomes discarded batches or illness risk. If a batch is unsafe but not recorded as discarded, the cost moves into a hidden public-health ledger. If service is delayed but reimbursement treats it as normal delivery, the fiscal record becomes less truthful.

This is why “demand response” is too narrow a term for the MBG kitchen question. The rupiah-relevant unit is not a kilowatt-hour alone. It is the operating consequence of moving or protecting that kilowatt-hour.

The public record that would make this measurable

A proportional public record would not need to expose private supplier data or turn every kitchen into a live dashboard. It would classify operating status in enough detail to distinguish safe flexibility from service risk.

A useful SPPG energy-status ledger would include:

• grid status and outage duration;
• generator availability, running hours, fuel stock, and fuel type;
• LPG or other cooking-fuel stock, cylinder size, and delivery interruption;
• refrigeration and freezer temperature status, with exception logs rather than continuous public temperature feeds;
• water source and pump status;
• critical-load exemption status during grid-stress events;
• flexible-load actions taken, such as delayed dishwashing, pre-cooling, battery charging, or water pumping;
• batches discarded, held, delayed, or reworked;
• dispatch delay by route window;
• whether an energy disruption changed reimbursement category.

The point is not to advocate a new bureaucracy. The point is to make an already public-scale operation legible enough that electricity flexibility does not hide food-safety costs and food-safety protection does not hide fuel costs.

This also fits the 3T question. Antara’s BGN report from March 2025 noted that 2,700 SPPG in 3T areas were being prioritized for gradual activation from September. Remote kitchens are less likely to have redundant energy, water, and logistics buffers. A single classification rule for urban and remote kitchens would be too blunt.

What the evidence does not support

The evidence does not support saying that MBG kitchens should broadly shed load when the grid is stressed. That would ignore refrigeration, water safety, cooking completion, ventilation, and dispatch timing.

The evidence also does not support saying that every kitchen load should be permanently exempt. Some loads can be scheduled. Some can be pre-cooled or pre-charged. Some can be delayed if clean inventory and hygiene rules are intact. Treating all kitchen electricity as sacred would waste a possible flexibility resource and may raise system costs for everyone.

The evidence does not support saying that VPPs defend the rupiah. At most, carefully measured flexibility can reduce some peak fuel burn, defer some expensive backup operation, improve visibility into imported-fuel exposure, and prevent hidden spoilage costs. Those effects are useful, but they are operating improvements, not currency protection by themselves.

The evidence does not yet show a mature Indonesian public framework for classifying food-service loads inside demand response. The public documents found for this piece show rooftop-solar quota rules, advanced-meter language, AMI discussions, kitchen construction and operating requirements, and food-safety outage principles. The bridge between them is still largely unwritten.

What remains uncertain

Three uncertainties matter most.

First, the actual load profile of SPPG kitchens is not public in a standardized form. We do not know how much electricity is used by refrigeration, pumps, ventilation, communications, dishwashing, preparation equipment, or lighting across kitchen types and regions.

Second, generator and LPG practice will vary. BGN’s technical guidance says backup power and gas infrastructure are required, but public reporting does not yet show how often generators run, how fuel is reimbursed, whether LPG stockouts are recorded as service risks, or how 12 kg and 50 kg cylinder availability differs across regions.

Third, Indonesia’s demand-response readiness is uneven. Advanced meters and rooftop-solar quotas make measurement more plausible, but they do not automatically create trusted curtailment classes, compensation rules, or food-safety exemptions.

The least-harm reading is therefore modest. Treat MBG kitchens neither as ordinary flexible load nor as untouchable load. Treat them as classified operating nodes. Flex the tasks that can move safely. Protect the functions that keep food, workers, and recipients safe. Record the rupiah costs when the boundary fails.