Technical Summary

On the evening of 31 March 2026, a fire at a Burundi National Defence Force (FDNB) ammunition depot adjacent to the Higher Institute for Military Cadres (ISCAM) in the Musaga suburb of Bujumbura initiated a catastrophic chain of explosions that continued for approximately six hours. Army spokesperson General Gaspard Baratuza attributed the ignition to an electrical short circuit. The depot contained a mixed inventory of artillery shells, grenades, and aerial bombs alongside non-explosive supplies including medicines, food, and mattresses. The resulting explosions killed at least 13 civilians and wounded 57 personnel (54 civilians and 3 soldiers). A further 8 inmates were killed at the nearby Mpimba central prison by falling debris and projectile fragments.

The reported effects are consistent with an Unplanned Explosion at a Munitions Site (UEMS) involving Hazard Division 1.1 mass-detonating ammunition. Projectiles were recovered up to 10 km from the depot. The base camp was described as “reduced to ashes.” These effects — sustained sympathetic detonation propagating through multiple ammunition natures over a six-hour period, with fragments and unfired projectiles dispersed across a wide area — indicate that the depot lacked both physical segregation between Potential Explosion Sites (PES) and adequate Quantity-Distance (QD) separation from the surrounding civilian population.

The location of this depot within a densely populated suburb of Bujumbura represents a fundamental violation of the Inhabited Building Distance (IBD) principles established in NATO AASTP-1 and the UN International Ammunition Technical Guidelines (IATG). The consequences were predictable and preventable.

Analysis of Effects

The six-hour duration of sustained explosions is the critical indicator of storage failure. A single detonation event — even of substantial NEQ — would produce a single blast event measured in milliseconds, not hours. A prolonged barrage of explosions over six hours indicates sequential sympathetic detonation: the initial fire or detonation propagated to adjacent ammunition stacks, which in turn initiated further stores. This chain reaction is characteristic of unsegregated bulk storage where different ammunition natures and Compatibility Groups are stored together without blast walls, traverses, or inter-magazine distances.

The mix of ammunition types reported — artillery shells (likely HD 1.1, CG D or F), grenades (HD 1.1, CG D with fuzes or HD 1.2 depending on configuration), and aerial bombs (HD 1.1, CG D) — should under AASTP-1 principles be stored in separate Potential Explosion Sites (PES) with calculated Inter-Magazine Distances (IMD) to prevent sympathetic detonation. The co-storage of these natures with non-explosive supplies (medicines, mattresses) indicates a general-purpose warehouse rather than a purpose-built explosives storehouse. This is not ammunition storage — it is warehousing, and the distinction is lethal.

The 10 km projectile throw distance is significant. Unfuzed artillery shells and aerial bombs projected by sympathetic detonation behave as Explosive Remnants of War (ERW) upon landing. Some may be intact but sensitised by thermal and shock exposure; others may have partially functioned fuzes. The entire debris field within the 10 km radius constitutes an ERW hazard requiring systematic clearance by qualified EOD personnel before civilian activity can safely resume.

Thermal Degradation and Climate Factors

Military analyst Egide Ruhashya correctly identified that tropical climate high temperatures can destabilise explosives. This is a well-documented phenomenon. TNT and Composition B fills in artillery shells undergo exudation and thermal cycling degradation when stored above 40°C without climate control. Propellant charges in uncontrolled tropical storage degrade through autocatalytic decomposition, producing nitric oxide gases that further accelerate decomposition. Without ammunition surveillance testing — periodic proof and inspection to assess energetic stability — there is no visibility of which lots have exceeded safe storage life. The absence of digital tracking or unique identification codes, as reported, means that even if surveillance were conducted, results could not be reliably linked to specific ammunition holdings.

Personnel and Safety Considerations

The siting of a military ammunition depot containing HD 1.1 mass-detonating stores within a densely populated suburb is the primary causal factor in the civilian casualties. AASTP-1 establishes Quantity-Distance tables that prescribe minimum separation distances between Potential Explosion Sites and Exposed Sites (ES) — including inhabited buildings, public traffic routes, and vulnerable facilities such as hospitals and prisons. For any meaningful quantity of HD 1.1 ammunition, the Inhabited Building Distance (IBD) would require separation measured in hundreds of metres to kilometres, depending on NEQ. The Musaga depot, surrounded by residential housing and adjacent to Mpimba central prison, could not have met any recognised QD standard.

The International Ammunition Technical Guidelines (IATG), published by the UN SaferGuard programme specifically for nations that are not NATO members, provide a scalable framework for ammunition storage safety. IATG Module 02.10 (Quantity and Separation Distances) and IATG Module 05.40 (Ammunition Storage Area Design) address precisely the failures observed at Musaga. These guidelines are freely available, technically sound, and specifically designed for implementation by national defence forces with limited infrastructure. Burundi’s apparent non-adoption of these guidelines — or any equivalent national standard — represents a governance failure with direct consequences for civilian safety.

Firefighting Response Limitations

Reports indicate that firefighters responding to the initial fire faced severe water shortages. This is operationally relevant: fires involving ammunition require specific response protocols. Under most national and NATO frameworks, fires in or near explosives storehouses trigger immediate evacuation to the applicable IBD rather than firefighting attempts, because the transition from fire to detonation may occur without warning. The decision to fight a fire in an ammunition depot must be weighed against the risk of personnel being within the lethal radius when sympathetic detonation occurs. Without knowing whether the responding firefighters were trained in explosives fire response protocols, this represents a further personnel safety concern.

Data Gaps

DATA GAP: Total Net Explosive Quantity (NEQ) — The total quantity of explosives stored at the depot has not been reported. The observed effects (10 km projectile throw, six-hour sustained detonation, complete destruction of the base camp) suggest bulk NEQ potentially measured in tonnes, but quantification is not possible without inventory data.

DATA GAP: Specific ammunition natures and quantities — “Artillery shells, grenades, and aerial bombs” is insufficient for technical analysis. The specific natures (calibre, type, fuze configuration), quantities, and lot numbers are unknown. Without these, precise HD/CG classification per STANAG 4123 cannot be assigned.

DATA GAP: Storage facility construction — Whether the depot comprised purpose-built explosives storehouses (earth-covered magazines, above-ground igloos, or field storage) or repurposed general buildings is not reported. The co-storage of ammunition with medicines and mattresses suggests the latter.

DATA GAP: Number of PES and inter-magazine distances — Whether the depot comprised a single storage building or multiple buildings with separation distances is not reported. The six-hour propagation timeline suggests multiple storage points that were sequentially initiated, but without site layout data this cannot be confirmed.

DATA GAP: IATG or national standard compliance — Whether Burundi has adopted IATG, any regional ammunition safety framework, or a national equivalent standard for explosives storage is not known from open sources.

DATA GAP: Ammunition condition and surveillance history — The age, condition, lot provenance, and surveillance testing history of the stored ammunition are not reported. The absence of digital tracking and unique identification codes (per Ruhashya) suggests no surveillance programme was in place.

DATA GAP: Electrical installation standards — Whether the storage facility electrical systems complied with any explosives-area electrical safety standard (intrinsically safe equipment, earthing/bonding, lightning protection) is not known. The official attribution to “electrical short circuit” raises the question of whether the electrical installation was suitable for an explosives environment.

DATA GAP: Fire detection and suppression — Whether any fire detection, alarm, or suppression systems were installed in the storage facility is not reported.