Aftershock Mode: The Operating Status MBG Kitchens Need During the Flores Sequence

MBG Watch · 2026-08-17

The premise

The Flores sequence is still active enough that MBG should not treat the question as a simple restart decision.

BMKG reported on 17 August 2026 that, by 11.00 WIB, it had recorded 1,624 aftershocks after the Flores tsunami-generating earthquake, with magnitudes from M1.3 to M6.2. BMKG said 23 aftershocks were above M5 and 59 had been felt. It described the main earthquake as a shallow M7.7 event on the Flores Back-Arc with a thrust-fault mechanism, and reported 53 deaths, 137 injuries, and damage across districts including Manggarai Barat, Manggarai, Manggarai Timur, Sikka, and Maumere, while urging people to avoid cracked or damaged buildings and remain alert for continuing aftershocks.

That matters for MBG because the program is no longer only serving ordinary school days in affected districts. RRI reported that the government had identified 71 Satuan Pelayanan Pemenuhan Gizi (SPPG) kitchens that could be used to support food needs for people displaced by the NTT earthquake. The same report quoted Home Affairs Minister Tito Karnavian saying some schools were still closed or had been closed because of damage, and that meals normally prepared for schools could temporarily be redirected to evacuation posts. It also reported BGN’s position that affected SPPGs should put safety first, report their condition, and adjust service under emergency conditions.

This is not evidence that MBG kitchens have failed. It is evidence that they are being asked to operate inside a changing disaster system: aftershocks, damaged schools, shelter feeding, uncertain routes, pressure on food logistics, and interruptions to energy and public services.

The operating question is therefore narrower and more practical than “continue MBG” or “stop MBG.” It is this: during a continuing shallow aftershock sequence, can the public see which kitchens are operating normally, which are in emergency-feeding mode, which are paused, and what conditions justify each status?

What is verified

First, the seismic hazard had not ended on 17 August. BMKG’s aftershock count, magnitude range, and public warning to avoid damaged buildings establish that the operating environment remained unstable. The tsunami warning had ended earlier, but the aftershock sequence had not.

Second, the food system around the response was already being widened beyond MBG. Antara reported that Bulog NTT added 10 tons of rice to strengthen food supplies for communities affected by the Flores earthquake, with deliveries prepared through El Tari Airport and Tenau Port and coordinated with BNPB for distribution to disaster posts. Bulog also said earlier rice and cooking-oil support had been sent through local BPBD offices to support public kitchens and evacuation posts.

Third, the route and energy environment was not ordinary. Antara reported that Pertamina added fuel-truck capacity from Bima to Flores to accelerate fuel normalization after the earthquake. It said 47 of 52 fuel stations in Flores had returned to operation, leaving five still not operating: two in Manggarai, two in Manggarai Timur, and one SPBUN in Sikka. The same report emphasized safety, infrastructure condition, and operational readiness as conditions for recovery.

For MBG, these facts connect directly to prior MBG Watch work. “When the Ground Moves Under the Kitchen” framed earthquake continuity as a food-safety issue, not only a disaster-management issue. “The Restart Ledger MBG Needs After the Flores Earthquake Escalation” argued that reopening requires a public record of building, equipment, water, power, staff, routes, and beneficiary conditions. “When MBG Kitchens Feed Shelters” and “Who Pays for the Shelter Meal?” separated emergency feeding from ordinary school-meal service and asked for distinct service and reimbursement ledgers. “The Route Is Part of the Kitchen,” “At the Kitchen Tap,” and “The Power Behind the Plate” each named one condition that becomes more important when a kitchen works under disruption: route integrity, water safety, and energy readiness.

The new crossing is that all of those ledgers now have to be visible at once.

What is not verified

I did not find, in the sources reviewed for this analysis, a kitchen-by-kitchen public operating-status table from BGN or local authorities showing which affected SPPGs were normal, restricted, paused, or converted to shelter feeding on 17 August.

I also did not find a public list tying each of the 71 possible SPPGs to named districts, named schools, named shelters, water status, power status, road status, batch-discard decisions, or safe-consumption time controls.

That absence should be stated carefully. It does not prove that the records do not exist internally. It does not prove that kitchens are unsafe. It does not prove that emergency feeding is being mismanaged.

It does mean the public cannot inspect the operating mode from the outside. In a disaster sequence, that matters. A kitchen can be safe at 06.00, constrained at 10.00, paused after an aftershock, and safely redirected by afternoon. A static “open” or “closed” label is too blunt for that reality.

What aftershock mode should record

Aftershock mode should be a public operating status, not a slogan. It should be temporary, practical, and inspectable. It should begin when a kitchen’s normal school-meal assumptions are disrupted by continuing seismic risk, shelter feeding, route instability, water interruption, power interruption, school closure, or local disaster-command instructions.

A minimal aftershock-mode record would have ten fields.

1. Kitchen status: normal, restricted, shelter-feeding, paused, inspection-required, or closed. This should be updated at least daily during emergency conditions, and more often after significant felt aftershocks.

2. Service status: which ordinary school beneficiaries are being served, which schools are closed, which meal allocations have been redirected, and which shelters are being served. A child not in school because the school is damaged should not disappear from the nutrition record.

3. Water status: source, continuity, treatment, storage, and the most recent basic safety check. “At the Kitchen Tap” argued for this in ordinary conditions; after a quake it becomes more important because pipe breaks, turbidity, and emergency storage can change quickly.

4. Power and fuel status: whether refrigeration, cooking, lighting, pumps, communications, and backup energy are reliable enough for the planned menu and volume. “The Power Behind the Plate” becomes a live control in this setting, especially when fuel normalization is still being reported in the wider Flores system.

5. Route status: which dispatch routes are open, delayed, rerouted, or unsafe; which schools or shelters are reachable; and which routes require shorter holding times. “The Route Is Part of the Kitchen” is not a metaphor during an aftershock sequence. If a road delay pushes a meal beyond the safe consumption window, the route has become part of the food-safety decision.

6. Safe-consumption time control: preparation time, dispatch time, arrival time, distribution time, and discard deadline. This is the field that prevents emergency generosity from becoming food-safety risk.

7. Menu substitutions: any changes made because cold-chain, water, fuel, market supply, or transport conditions have changed. Substitutions are not a problem by themselves. Unrecorded substitutions are.

8. Discarded batches: what was discarded, why, where, and when. This protects children and protects kitchens from pressure to distribute food that has become unsafe.

9. Incident reporting: illness complaints, damaged facilities, near misses, aftershock interruptions, road failures, water failures, and power failures. The record should include “none reported” when none are reported, not only publish bad news.

10. Return-to-ordinary trigger: the conditions that must be true before the kitchen exits aftershock mode. These should include building and equipment check, water and power stability, route reliability, school/shelter service clarity, and no unresolved safety incident.

The point is not to burden a disaster response with paperwork for its own sake. The point is to reduce confusion. In an emergency, a short public status record can prevent three harms at once: unsafe meals, missed beneficiaries, and opaque spending.

The fiscal and infrastructure crossing

The sister-organization crossing is real in principle, but should be handled soberly. Rupiah Stability Watch’s frame — emergency feeding as potentially currency-relevant infrastructure — should not be read as saying one damaged kitchen moves the rupiah. That would be too large a claim.

The narrower claim is stronger. When a national social program becomes part of disaster logistics, it touches fiscal credibility and public-infrastructure confidence. If SPPG kitchens are used to feed shelters, someone must record which meals were ordinary MBG obligations, which were emergency meals, which costs should be reimbursed, and which infrastructure failures changed the cost of delivery. “Who Pays for the Shelter Meal?” made that budget point. Aftershock mode adds the operating point: reimbursement without a status ledger is hard to audit, and a status ledger without reimbursement fields is hard to sustain.

This is where public finance meets food safety. A discarded batch is a safety action, but it is also a cost. A fuel delay is a route problem, but it is also a service interruption. A school closure changes where a child receives food, but it also changes which ledger the meal belongs to. None of these facts needs blame attached to it. They need timestamps.

The least-harm path

The least-harm path is not a blanket pause. It is also not a blanket continuation.

A kitchen with intact structure, safe water, reliable power, reachable service points, trained staff, and a clear safe-consumption clock may be able to operate normally or support nearby shelters. A kitchen with damaged walls, uncertain water, unstable refrigeration, blocked roads, or unresolved illness complaints should pause or restrict service until the condition is corrected. A kitchen near affected areas but not directly damaged may be useful as a backup producer, but only if its route and holding-time controls remain safe.

The public record should allow each of those decisions to be seen without turning every operational adjustment into a scandal.

For the next phase of the Flores response, MBG and local disaster authorities should publish a daily aftershock-mode table for affected districts. It can be simple: kitchen name or code, district, current status, service destination, water status, power status, route status, safe-consumption control, discarded batches, incidents, reimbursement category, and return-to-ordinary trigger.

That table would not decide whether MBG is good or bad. It would make the actual decision visible: which children and displaced families were fed, by which kitchen, under what constraints, and with what safety controls.

What I am uncertain about

The main uncertainty is the internal record. BGN, SPPG operators, BPBD offices, and local governments may have operational lists that are not yet public or not easily discoverable. If those lists exist, the recommendation is not to create a new system from nothing; it is to publish the public-facing subset that communities, schools, shelter managers, auditors, and health officials need.

The second uncertainty is the district-level condition of individual kitchens. The sources reviewed establish continuing aftershocks, casualties, school disruption, emergency-feeding intent, food-logistics support, and fuel-normalization work. They do not establish the status of named SPPG kitchens.

The third uncertainty is timing. Aftershock mode should end when operating assumptions have returned to ordinary conditions, not on a symbolic date. BMKG’s public updates are one input, but the kitchen-level trigger has to be local: building safe, water safe, power reliable, route workable, school or shelter destination clear, and no unresolved food-safety incident.

That is the practical standard: not stop, not continue, but show the mode and the reason.