When a Whistle in the Wind is the Sound of Steam: Lessons Learned from a Building Emergency

Susan Leite

How do you recall the evening of Friday, October 31, 2008? For most, the evening held some connection with Halloween: that occasion known for flights of fear or pranks of all scales. For the occupants of Building 66, however, that Halloween Friday will be remembered as a trick in the form of a steam release and a lesson learned in the unpredictability of emergencies and the necessity of a business continuity strategy.

At 8:30 pm, condensation-induced water hammer occurred in 200 psi steam piping in the Building 66 sub-basement mechanical room. The hammer forces caused pipe anchors to pull out of the concrete ceiling slab, allowing enough pipe movement for an expansion joint to pull apart.

The Expansion Joint Pulled Apart
(click on image to enlarge)

From the two open pipe ends a significant volume of steam was released into the building. Thirty sprinkler heads fused due to the heat of the steam. Sprinkler water and condensed steam caused significant damage in the building. Subsequent inspection the following morning revealed the failure of a steam trap and damage observation supports the conclusion that the condensation-induced water hammer occurred at that location. Fortunately, no injuries or fatalities occurred and Department of Facilities staff worked ceaselessly over a 24-hour period to maintain essential building services and to ensure re-occupancy the following Monday.

In the following weeks, MIT engaged two independent third-party investigators to address the root cause of the incident and identify any follow-up recommendations. Professor Emeritus Peter Griffith of MIT’s Mechanical Engineering Department is considered an authority on heat transfer and thermodynamics and was able to lend his expertise to the assessment of the incident. From these reports, the Department of Facilities identified preventive measures that would mitigate the likelihood and impact of a recurrence.

Mitigating Factors

Several factors mitigated the impact of the steam line rupture. First, the timing of the release, combined with the evening being Halloween, meant the building was minimally occupied and no one was in the direct path of the release. Second, surveillance video revealed that approximately 90 seconds elapsed before steam infiltrated through hallways and doors. Those still in the building and immediately responding to the fire alarm had time to safely evacuate. Some air-handling units remained functional, which mitigated temperature effects. Areas with the most water damage either did not contain water reactive chemicals, or the containers were properly stored under inert atmospheres. Finally, the Facilities Manager and Environment, Health, and Safety (EHS) staff surveyed the building within 24 hours of the incident; this was essential to identifying any operations needed to make the building safe for re-occupancy.

Recovery

The recovery plan for Building 66 is a collaboration engaging the Department of Facilities, the Chemical Engineering Department Facilities Manager, the MIT Insurance Office Claims Adjuster, FM Global (MIT’s property insurance carrier), the EHS Office, and the Security and Emergency Management Office. The Chemical Engineering Department was fortunate in that no laboratory spaces experienced a complete loss of research facilities. Loss consultants who surveyed the extent of equipment and facilities damage applied the following guidelines for estimating individual lab losses:

• collect purchase/rental agreement records for all equipment acquired or leased for the lab.
• inventory all equipment with disposition status: salvage, test, or toss.
• if safe to do so, plug in equipment to determine if immediately functioning.
• equipment with sensitive electronics could be vulnerable to a steam condensation effect that may be beyond the temperature tolerance of the electronics; this equipment could require replacement.
• equipment with motor housings submerged in water could likely fail at a future time.
• request that equipment vendors test equipment on site, rather than sending equipment off site.
• review agreements for any equipment on loan and test and salvage equipment accordingly.
• plan on a lead time of at least eight weeks for ordering equipment replacements.

  In December 2008, the Department of Facilities and the Chemical Engineering Department entered into a scope of work for reconstruction and repair of the building facilities, with the rebuild schedule slated to occur over several months.

  Back to top

  The Value of Multi-Hazard Risk Identification, Assessment, and Mitigation

  Though the incidence of steam line releases at MIT is extremely low, they are potentially high consequence events in terms of injury/fatality potential, loss of building function, and secondary hazards. What this spectrum of emergency scenarios that can impact a decentralized organization like the university setting underscores is the importance of a method to identify potential risks using a multi-hazard approach, and the need to develop strategies for their mitigation, including an effective business continuity strategy across research and administrative units.

Efforts to identify and mitigate impacts from a spectrum of man-made and natural hazard scenarios were the focus of a research project funded under the Federal Emergency Management Agency’s Disaster Resistant University (DRU) program. The project, under the supervision of Professor George Apostolakis, who has dual faculty appointments in the Department of Nuclear Science and Engineering and the Engineering Systems Division, entailed developing a quantitative framework for identifying and assessing a universe of risks and their impact (or disutility) through a variety of lenses such as human and/or environmental health, infrastructure resilience, intellectual property, and reputation.

The research effort engaged numerous stakeholders representing a cross section of the Institute’s academic, research, and administrative units and generated a dialogue on the following questions and their role in building organizational resilience:

• for a given hazard scenario (e.g., flood, fire ) how frequently does such an incident occur on our campus or is reported elsewhere? if losses are incurred, what is the value?
• if an outage occurs, what is the typical duration?
• if a particular operation experiences an outage, what is the critical timeframe for restoring operations?
• what mechanisms are in place to mitigate the risk, and what steps can be taken to enhance resilience moving forward?

Finally, researchers evaluated interdependencies across campus – for departments, this is where the proverbial rubber hits the road – understanding that the research and education enterprise is a network of building infrastructure, communications, and utilities systems, locally operated and maintained research equipment, administrative services, and the numerous departments which provide or use the service of a core facility that may support hundreds of research collaborations and represents millions of dollars of research investment.

Building organizational resilience is an effort that commands involvement across all departments and cannot be effectively achieved through delegation or outsourcing. The all-hazards assessment framework is a valuable heuristic technique for evaluating low probability-high consequence events alongside higher probability-lower consequence events. Lessons learned from the steam incident and events elsewhere continuously refine the model, enriching its future value as a decision-making tool and model for emergency planning.

Empowering Your Own Organizational Resilience

In the December 2007 MIT Faculty Newsletter, an article on the lessons learned from the December 2006 fire at One Broadway which resulted in an extended outage for several departments, summarized key considerations for emergency planning that are applicable to a whole host of scenarios including the steam incident described earlier. In underscoring the importance of proactive extended outage and emergency planning, academic, research, and administrative departments across the Institute are asked to consider the following:

• how would you stabilize research activity if water service or power was lost for more than a day or a week; your building was closed due to a major fire, gas leak or other building system emergency for an extended period of time; or a significant natural disaster renders your research area indefinitely inaccessible?
  
  
  • what are the major supply chains for your lab operations? What contracts are currently in place with vendors?
    
    
  • is research and/or office data routinely backed up off site?
    
    
    • has an alternate location been identified which includes the resources necessary to continue your group’s business operations on a temporary basis?
      
      
    • have you discussed how to handle emergency or outage with staff and students?
      
      
      • do you have ready access to department phone lists and MIT office contacts that can assist you in an emergency? is there a strategy for communicating if e-mail and cell phone services are down?
        
        
      • are building occupants well-versed in evacuation routes and rallying points? does the plan address ADA (Americans with Disabilities Act) needs?
        
        
        • have individuals updated contact information on MIT Alert in the event of a campus emergency?

        All units are well-advised to engage in the process of identifying their critical activities and plan for the eventuality of an extended outage (e.g., implementing a local communications plan as well as ensuring support for activity that cannot be discontinued during an emergency). Bill VanSchalkwyk, Managing Director of Environmental Health and Safety (EHS) Programs, has made it an EHS priority to assist departments with developing local emergency preparedness and business continuity plans. Together, we can ensure a safe and orderly response to an emergency or outage and protect MIT’s most important assets: its people and its research.

	Back to top
	Send your comments