Skip to content

Isotec/Sigma Aldrich Nitric Oxide Explosion

Overview

Explosion at a biochemical facility involving a liquid nitric oxide release at Isotec, Miami Township, Ohio, on September 21, 2003. One employee was injured, and more than 2,000 local residents were evacuated for 24 hours.

Incident Snapshot

Field Value
Facility / Company Isotec, a wholly owned subsidiary of Sigma–Aldrich Corporation
Location Miami Township, OH
Incident Date 09/21/2003
Investigation Status CSB issued a case study report on this accident on August 24, 2004.
Accident Type Chemical Manufacturing - Fire and Explosion Investigation Status: CSB issued a case study report on this accident on August 24, 2004.
Final Report Release Date 08/24/2004

What Happened

  • At about 7:30 am, the Isotec on-call system operator received an automatic pager alert indicating an alarm condition in a cryogenic nitric oxide (NO) distillation unit.
  • Arriving at the facility at about 7:50 am, he observed reddish-brown gas venting from the distillation unit vacuum pump exhaust, indicating a breach in the column piping within the vacuum jacket.
  • The responding employee immediately notified his supervisor, who called the 911 dispatcher.
  • By 8:15 am, employees secured the leak by closing the vacuum pump suction valve.
  • Personnel began installing temporary tubing to empty the nitric oxide in the malfunctioning column and closely monitored the pressure inside the column.
  • The pressure stabilized at no more than 130 psi.
  • At 10:15 am, without warning, a violent explosion destroyed the distillation column, the blast containment structure, and nearby buildings.
  • Immediately following the explosion, the incident commander ordered an evacuation of the community within a 1-mile radius.
  • The evacuation order was lifted after 24 hours.

Facility and Process Context

  • Isotec is a wholly owned subsidiary of Sigma–Aldrich Corporation.
  • Sigma–Aldrich purchased the Isotec facility in 2001 from Matheson Gas Products.
  • The facility is located in Miami Township, Ohio, 12 miles south of Dayton and 0.5 mile west of Interstate 75, on an 11-acre parcel.
  • Over the past two decades, adjacent land use changed from rural farmland to established residential development; more than 500 homes are located in the surrounding area.
  • Isotec employs about 75 full-time personnel working on an 8-hour, 5-day schedule.
  • Ten employees are involved in distillation unit operations.
  • Two NO distillation units—N3 and N6—were operating at the time of the explosion.
  • Two other NO distillation units were installed but not operational.
  • The NO process operates 24 hours a day, 7 days a week.

Consequences

  • Fatalities: 0
  • Injuries: 1 employee injured; glass shards lacerated the hand of an Isotec employee.
  • Environmental release: Nitric oxide vented to the atmosphere and reacted with air to form nitrogen dioxide; a ruptured fill line vented CO gas, which then ignited and burned for about 1 hour.
  • Facility damage: The explosion destroyed the distillation column, the blast containment structure, and nearby buildings; windows were blown out of the main office building; small chunks of concrete and metal shards were propelled as far as 1,000 feet; a 52,000-pound gaseous carbon monoxide storage vessel was pushed about 10 feet off its foundation; a second steel panel severely damaged adjacent equipment; the column, associated equipment, adjacent buildings, and blast containment structure were destroyed.
  • Operational impact: More than 2,000 local residents were evacuated for 24 hours; Isotec/Sigma–Aldrich informed CSB and the community that they would not restart NO distillation and would decommission the remaining NO unit, N6.

Key Findings

Immediate Causes

  • A leak developed in the distillation column piping, releasing nitric oxide into the vacuum jacket.
  • Nitric oxide continued entering the vacuum jacket through the leak in the pipe and detonated, crushing the column piping and bursting the jacket.

Contributing Factors

  • Loss of vacuum in the jacket seriously degraded its insulating capacity, thus increasing the heat load on the column.
  • Closing the vacuum jacket isolation valve to stop the NO leak into the environment pressurized the vessel as the liquid nitric oxide began to boil.
  • A shock load of sufficient energy to detonate liquid nitric oxide could result from rapid vaporization of liquid nitric oxide as it entered the vacuum jacket.
  • A shock load of sufficient energy to detonate liquid nitric oxide could result from collapse of a reboiler from pressure buildup in the vacuum jacket.
  • The rectangular blast shield structure directed a significant quantity of small debris vertically up, most likely preventing serious injury to the employees and emergency responders in the vicinity.
  • The total destruction of the blast shield structure caused major damage to the surrounding facility, including dislodging the CO vessel and rupturing the CO piping.
  • The Isotec structure omitted a heavy steel wire mesh top to capture debris.
  • The Isotec structure omitted a wide top-to-bottom labyrinth opening in one wall to prevent pressure buildup.
  • No pressure relief device was installed on the vacuum jacket.
  • The jacket was equipped with a block valve which, when closed, could result in significant pressurization.

Organizational and Systemic Factors

  • The Isotec/Sigma–Aldrich process hazard analysis (PHA) team acknowledged that liquid nitric oxide presented an explosion hazard; however, the team did not understand the significance of the risk to employees.
  • Although at least two PHAs documented that detonation of liquid nitric oxide is a "credible scenario," neither analysis comprehensively addressed the previous incidents involving NO detonation.
  • The PHAs did not thoroughly review administrative and engineering controls or the consequences of postulated and actual failures.
  • There was no system in place to track PHA findings and associated followup actions.
  • Isotec management did not adequately investigate the two previous NO detonation incidents involving the other distillation units.
  • Because the specific causes of the N4 column failure and NO release in 1998 were not determined, no appropriate corrective actions were implemented to prevent a similar failure in the N3 column.
  • There was no record of actions taken to apply lessons learned from the N4 incident, or the failure of column N2 in 1995, to the design and operation of the N3 distillation unit.
  • Neither the township zoning process nor the township and city permit approval processes adequately considered the hazards of preexisting industrial chemicals.
  • Neither authority prescribed steps for addressing potential public consequences from the accidental release of chemicals.

Failed Safeguards or Barrier Breakdowns

  • The NO distillation units incorporated NO detectors and alarm systems, but the incident still escalated to explosion.
  • Temperature and vacuum jacket pressure alarms were integrated into an automatic dialing pager system to alert key personnel of a process upset, but the explosion occurred without warning.
  • The automatic shutdown signal caused the column to go into full reflux, but this did not prevent the explosion.
  • Passive protective devices were the primary means of minimizing the effects of an NO explosion, but the blast shield structure was not fully effective.
  • The exterior blast containment structure omitted a heavy steel wire mesh top.
  • The exterior blast containment structure omitted a wide top-to-bottom labyrinth opening in one wall.
  • No pressure relief device was installed on the vacuum jacket.

Recommendations

  1. Recommendation ID: Not provided
    Recipient: Not provided
    Status: Not provided
    Summary: No formal recommendations were provided in the source extract. The recommendations array retains the required placeholder.

Key Engineering Lessons

  • Liquid nitric oxide is an unpredictable, highly shock-sensitive explosive and must be treated as such in process design and hazard analysis.
  • Vacuum jacket design should account for loss of vacuum, pressurization during isolation, and the potential for detonation from rapid vaporization or pressure-induced collapse.
  • Passive blast containment structures should include features identified in the report as more effective, including a heavy steel wire mesh top and a wide top-to-bottom labyrinth opening, to better control debris and pressure effects.
  • A vacuum jacket without a pressure relief device and with a block valve that can significantly pressurize the vessel presents a serious design vulnerability.
  • Previous NO detonation incidents must be investigated thoroughly and translated into corrective actions for similar units.

Source Notes

  • Consolidated from the CSB final report extract (source_priority 1).
  • No formal recommendations were provided in the source extract; the recommendations array retains the required placeholder.
  • All facts were taken only from the provided source text; no external information was added.

Similar Incidents

Incidents sharing the same equipment, root causes, or hazard types.

Same Equipment

Same Root Cause

Same Hazard

← View in Knowledge Graph