Case Study: In 2007, one of our clients indicated that they expected that they could run one 750-ton chiller on their EPS while providing all of the other life safety, critical, and equipment loads. We decided to test this during a normal scheduled outage at 2:00am in May 2007. We found that given the loads at that time, we could actually run two 750-ton chillers and associated pumps.
In August 2007, the local utility decided to conduct maintenance on the substation serving the hospital. To accomplish this, they brought in a trailer-mounted substation to provide power during the maintenance outage. The hospital was not informed of this activity. During the maintenance work, the trailer-mounted substation failed. The EPS at our clients hospital started normally and was providing life safety, critical, and equipment branch power. Virtually all of the air-handling units (AHUs) started automatically on the equipment branch, but chillers are started manually. Secondary chilled water pumps are also on the equipment branch. This outage started around 1:00pm on a hot, humid August Monday. All HVAC controls were set to operate normally. Thus, as the secondary loop chilled water temperature rose, all chilled water coils went to full, open position. Outside air dampers remained in their normal position.
Within 30 minutes, temperature and humidity were noticeably rising. Shortly thereafter, condensation began to form on walls and floors, creating among other problems, slippage hazards. At this point, engineering decided to manually start one of the 750-ton chillers. Unfortunately at this time, the estimated cooling load was approximately 2,500-tons. Thus, chilled water temperature continued to rise. After about one hour, the engineering department, remembering our test from May, decided to start the second 750-ton chiller. Needless to say the load on a hot, August afternoon was far greater than the load during the May 2:00am test. Thus when the second 750-ton chiller was started, the in-rush current brought down the entire EPS and the 1million square foot hospital went black. To make matters much worst, the lead electrician was on vacation and the methods for restarting the EPS were complex and not well understood by others. Quick work by the remaining staff with the help of the former lead electrician, now head of disaster management, limited the total blackness to approximately 15 minutes. For the next four hours, the hospital ran only one chiller and temperatures rose into the mid 80s. Power was restored around 6:00pm. Luckily, there were no serious adverse patient outcomes as a result. Shortly thereafter, our firm was engaged to develop a plan for operation under an EPS.