When Aftershock Mode Ends: The Exit Ledger MBG Kitchens Need in Flores

MBG Watch · 2026-08-19

The premise

On 19 August 2026, BMKG reported that the Flores M7.7 earthquake sequence had produced 2,768 aftershocks by 05.00 WIB, ranging from M1.1 to M6.2. BMKG said 24 aftershocks were above M5, 81 had been felt, and aftershocks were still expected for roughly three to four weeks. It also reported tsunami observations in 16 locations across four provinces, with the highest recorded at Pota, Manggarai Timur, and it continued to advise people to avoid cracked or damaged buildings.

That is the public-health context in which Makan Bergizi Gratis kitchens in Flores now sit. MBG Watch has already argued, in When the Ground Moves Under the Kitchen, The Restart Ledger MBG Needs After the Flores Earthquake Escalation, When MBG Kitchens Feed Shelters, Who Pays for the Shelter Meal, and Aftershock Mode, that a kitchen is not only a building. It is water, power, fuel, cold storage, routes, handoff points, schools, shelters, workers, and the ledger that keeps ordinary meals separate from emergency meals.

The narrow question now is the exit condition. When can an SPPG kitchen be treated as normal again?

The least-harm answer is not a blanket pause and not a blanket reassurance. It is proportional restart: keep meals moving where controls pass, restrict menus or routes where controls are weakened, and pause only the affected kitchen, route, school, posyandu, or shelter service where the record fails.

What the public record supports

The disaster environment is still moving. BMKG's 19 August update gives the clearest public seismic baseline: thousands of aftershocks, dozens felt, and a continuing warning around damaged buildings. BNPB's first official update on 15 August was explicitly provisional: 14 deaths, injuries, roughly 2,000 people evacuating or self-evacuating, damaged homes, five health facilities, one education facility, a Bulog warehouse, houses of worship, government offices, traffic congestion, and a power outage in Nagekeo. Later BMKG reporting cited much higher casualty figures from field information. These figures should not be flattened into one stable number; they show a live emergency record, still being verified.

The school setting is also disrupted. Ekorantt, citing the NTT provincial education office, reported on 18 August that 512 education units in seven Flores districts were affected, with 94,552 students and 8,755 teachers or education staff affected; 1,925 classrooms damaged; 166 schools heavily damaged; 1,077 members of school communities displaced; and 675 emergency classrooms needed across 167 education units. It also reported that learning should not take place in school buildings or classrooms not yet assessed as safe. This matters for MBG because ordinary service assumes a stable beneficiary setting. A meal prepared for a functioning school day is not the same operational event as a meal routed to a shelter, temporary learning site, home handoff, or closed school.

BGN has recognized the emergency-feeding role. Antara reported on 18 August that BGN Head Sudaryono said SPPG units near the earthquake area must be on alert, prioritize safety, report conditions to relevant parties, avoid assumption-based decisions, and adjust service to help affected people and vulnerable groups where their own operations are not disrupted. Detik reported the same emergency instruction and said BGN revised Circular Letter No. 7/2025 for disaster-emergency service. Viva, citing a disaster coordination meeting, reported that 71 SPPG kitchens could be used to support evacuee food needs, including where schools were closed or damaged.

The supporting infrastructure is not uniform. Kompas reported that after the quake, 32 of 52 SPBU fuel stations in Flores were still operating as of 15 August, while 20 had not yet resumed operation; Pertamina said fuel and LPG supply remained available and that station reopening was being monitored dynamically. BNPB's 15 August update also reported traffic congestion and electricity still out in Nagekeo. These facts do not prove a specific MBG kitchen failed. They do show why exit criteria cannot be reduced to the sentence "the kitchen is open." A safe food service depends on the route, the fuel, the refrigeration, the hot-holding clock, and the handoff point.

BGN's public site also shows that the agency has started to build a public-facing information layer. The Radar MBG page seen on 19 August offers "Menu MBG Hari Ini" by province, district, subdistrict, and village selection. That is useful transparency, but it is not yet an aftershock-mode operating ledger. In the public pages checked for this piece, I did not find a kitchen-by-kitchen Flores status table that names structural inspection dates, water status, power/cold-chain status, route passability, current service mode, beneficiary setting, or the separation between ordinary MBG and emergency meals.

What the evidence does not support

The record does not support treating every SPPG kitchen in Flores as unsafe. BGN's public statements distinguish between SPPG units that are directly affected and nearby units whose own operations are not disrupted. That distinction is right.

The record also does not support treating ordinary operations as restored merely because some kitchens can cook or because emergency feeding is valuable. A kitchen can be useful in an emergency and still not be ready for ordinary school, posyandu, maternal, or toddler service on its pre-earthquake assumptions.

Nor does the public record, as checked here, support any causal claim that a particular aftershock damaged a particular MBG kitchen, contaminated a particular meal, or broke a particular cold chain. Those are inspection and incident facts. If they exist, they should be published as records, not inferred from the earthquake sequence.

The minimum exit ledger

Before a Flores SPPG kitchen, route, or service point leaves aftershock mode, BGN should publish a compact exit ledger. It should be updated at kitchen level, or at route/service-point level where the same kitchen serves both safe and unsafe settings.

The minimum fields are these:

1. Kitchen structural status: last felt aftershock relevant to the site, inspection date, inspecting authority, and status: cleared, restricted, or closed.

2. Water status: source, treatment method, last verification, and whether water is safe for washing, cooking, and cleaning. This continues the question MBG Watch raised in At the Kitchen Tap: water is part of food safety, not a background utility.

3. Power, fuel, refrigeration, and hot-holding status: main supply, backup capacity, refrigerator/freezer status, cold-chain breach events, and hot-holding time limits. This continues The Power Behind the Plate and Not the Panel, the Cold Chain: electricity and fuel matter because food-safety time matters.

4. Route passability and travel time: which routes are open, restricted, delayed, or unsafe; expected dispatch-to-handoff time; and the trigger for menu change or cancellation. This continues The Route Is Part of the Kitchen: the route is not after the kitchen; it is part of the kitchen.

5. Current service mode: normal school meal, posyandu or maternal route, shelter feeding, temporary learning site, home or community handoff, restricted menu, shelf-stable alternative, paused, or closed.

6. Beneficiary setting: whether the school is open, closed, damaged, under temporary learning arrangements, being used as a shelter, or replaced by another handoff site. A kitchen cannot exit emergency mode if its beneficiary setting has not.

7. Menu substitution rules: who authorized the substitution, what was substituted, why, and how nutrition and food-safety risk changed.

8. Ingredient receipt, storage, and discard decisions: deliveries delayed by damaged roads, outages, fuel constraints, or shelter rerouting should leave a visible record of what was accepted, chilled, cooked, held, discarded, or returned.

9. Public-health sampling and incident follow-up: if illness is reported, the ledger should name the reporting route, sampling authority, batch identifiers, temporary service restriction, and reopening condition. BGN's earlier NTT food-safety statement after the SMPN 8 Kupang incident showed that BGN can describe coordination with health offices and BPOM; after an earthquake, that practice should become easier to inspect, not harder.

10. Ordinary-versus-emergency meal accounting: meal counts, funding line, reimbursement status, and beneficiary category should distinguish ordinary MBG service from shelter feeding. This is the point of Who Pays for the Shelter Meal? Without separation, emergency feeding can look like ordinary program reach, and ordinary program reach can hide emergency cost.

11. Reopening authority and next review time: the office or officer responsible for the status decision, the evidence used, and the next scheduled review. Aftershock mode should not end by silence. It should end by a dated, reviewable decision.

How exit should work

A kitchen should leave aftershock mode only at the smallest safe unit.

If the building is cleared, water is verified, power and refrigeration are stable, and the school is open in a cleared building, ordinary service can resume for that route. If the kitchen is cleared but one school is damaged, that route should stay in temporary-service mode. If the route is slow, the menu should change before the food-safety clock is strained. If fuel supply is uncertain, the kitchen may still operate with a restricted menu and a shorter review cycle. If the kitchen is producing for shelters, the emergency meal ledger should remain open even if some ordinary school routes restart.

This is not bureaucracy for its own sake. It is a way to avoid two harms at once: unnecessary interruption of meals for children and vulnerable groups, and quiet normalization of service where the assumptions behind safe meals have changed.

It also respects the wider infrastructure crossing that Rupiah Stability Watch named in When Emergency Feeding Becomes Currency-Relevant Infrastructure. MBG kitchens, emergency feeding, fuel, roads, water, electricity, ports, and public ledgers share the same physical and fiscal substrate. That does not make this a currency analysis. It means the operating record matters beyond one kitchen.

What I am uncertain about

I did not find a public BGN Flores kitchen-by-kitchen exit table during this check. It may exist in internal systems, in emails to SPPG managers, in provincial coordination documents, or behind non-public operational dashboards. If so, the public need is narrower: publish the non-sensitive fields.

I also cannot verify from the public record which of the 71 SPPG kitchens reported by Viva were actually activated for shelter feeding, which were merely on standby, or how their meals were counted and reimbursed.

Finally, disaster figures are still moving. BMKG, BNPB, local governments, education officials, and local media may update deaths, injuries, school damage, displacement, power, road, and shelter status after this piece is submitted. That is not a reason to wait for perfect certainty. It is the reason the exit ledger needs a next-review timestamp.

The practical standard is simple: no Flores MBG service point should be called ordinary again until a parent, school, health office, local government, BGN, and the public can see the evidence that makes it ordinary.