George Harrell is associate vice chancellor for administration and finance - facilities at East Carolina University, Greenville, North Carolina. ECU was a 1999 recipient of the Award for Excellence in Facilities Management, APPA's highest institutional honor. He can be reached at harrellge@mail.ecu.edu..
On Thursday, September 16, 1999, Hurricane Floyd came ashore and dropped more than 20 inches of rain. The flooding that resulted was far worse than would be expected from that amount of rain. The "Flood of the Century" (actually a 500-year flood event) was the result of the double hit by Hurricane Dennis two weeks earlier, that soaked the area with 12" of rainfall, and Hurricane Floyd. The combined effect was 32 inches of rainfall in a two-week period, half our expected annual rainfall.
East Carolina University's main campus has 3 million square feet of buildings located on 400 acres near the center of Greenville, North Carolina. We were a few weeks into the fall semester, the busiest time of the year, with a record enrollment, when Hurricane Floyd struck. It was the middle of the week and there were over 4,000 students on campus in residence halls.
A few weeks earlier, Hurricanes Dennis "One" and Dennis "Two" had not been much more than an annoying amount of rain; no one at the time could imagine how this would affect future events. When Hurricane Floyd hit, the initial flash flood produced most of the physical damage to the campus. We experienced almost none of the typical wind damage usually associated with a hurricane. In fact, not a single window was broken, and only one roof structure was breached.
We did lose several very large trees because the super saturated soil, combined with Floyd's winds, resulted in entire root bundles (some 20 feet in diameter) being pulled out of the ground. Normally, a knee-deep stream called Green Mill Run meanders through the campus. It carries storm water run-off from the campus and the City of Greenville. That small stream turned into a river 300 yards wide flowing at 6-8 knots.
The flash flood that occurred within a two-hour period inundated two of the largest and most critical classroom buildings on campus with 54 inches of water. The damage was over $1.5 million. The lower levels of the two buildings house the mechanical rooms that have chillers, air handlers, transformers, electrical switch gear, fire safety panels, and fiber optic nodes. In other words, all the expensive systems in the building. One of the buildings had the primary steam distribution and condensation collection center for most of the campus and a major fiber optic network node controller valued at approximately $100,000.
The flooded buildings and high voltage electrical distribution system damage was obvious. What initially escaped our attention was the underground steam distribution system. The water flooded the system and soaked the insulation, the water turned to steam and literally blew everything apart except the actual steam pipe. This resulted in $2.1 million of damage and has required total replacement of the primary steam feed for the campus.
We assessed our damage and started to develop a plan of action that would ensure the safety and well being of students, get rid of the water, sanitize and initiate emergency repairs, reopen two academically critical campus buildings, and repair the utility infrastructure. All of this needed to be accomplished within a two-week period or we faced the possibility of losing the entire semester.
This doesn't sound too bad until you look at the complications caused by the widespread devastation to the region. The flooding had washed out literally hundreds of roadways. For several days, ECU and the City of Greenville had no passable roadways in or out of the city. It was almost impossible to get materials, supplies, or services. Most businesses in the region were closed, including the contractor support that we needed to assist in our recovery efforts.
Many of our dedicated employees were either flooded out, unable to find a passable route into the city, or marooned north of the Tar River that now cut Greenville in half. We found ourselves with about half of our normal trades workforce and one-fourth of our housekeepers. We later determined that nearly 25 percent of our facilities workforce had lost everything-homes, cars, and personal possessions.
In addition, several of our senior managers and key members of our "engineering brain trust" were stranded at the APPA Institute for Facilities Management in Montreal, Canada. They couldn't fly back to Greenville because the airport was under water, and if they flew elsewhere, there was no ground route into Greenville. Campus food supplies were sufficient for only a few days due to closed roads. We were feeding 4,000 students, our own critical personnel, and many other emergency personnel in town. We had the only dining facilities in operation in the city.
We had a reduced workforce, full residence halls, over taxed dining facilities, food supplies running short, and an extraordinary amount of repairs to accomplish in a relatively short period. We also had a nationally televised football game with the University of Miami Hurricanes scheduled for Saturday. A home football game would add to the strain on the campus workforce and resources.
Our primary responsibilities were the safety and well being of our students and reopening the university. Continuing to house and feed students and staging a football game would consume resources that needed to be redirected to reopening the campus for business. If we could not reopen the two large academic buildings (one a major laboratory building and the other the largest classroom building on campus) within two weeks, we were faced with canceling the entire semester with tremendous short-term and long-term implications. We had to concentrate on our primary objective and direct all our limited resources to reopening.
During natural disasters and other emergencies, the campus is managed via the Campus Emergency Response Team. This is a relatively small group that cuts across the bureaucratic structure of the campus. This group is convened before an event with a preparation objective, and then during and after an event with a recovery objective. While our preparation for Hurricane Floyd was fairly routine, how we dealt with the aftermath of the event was anything but routine.
The Emergency Response Team was faced with having to make several major decisions very quickly, these included:
It was most fortunate that we made these decisions, because things only became worse. The flash flood was followed by the "Flood of the Century." Roads flooded (1,200 or more were breached) as the Tar River crested at an all-time high. Luckily, the campus is high enough not to reflood the buildings; however, it completely cut off everyone, including our workforce, living north of the Tar River.
Add to this, the only electrical substation feeding the entire city was located in the flood plain, and two of the primary transformers had already been shut down as the water level increased. The third and final transformer had to be temporarily deenergized to bypass safety switch gear that was going under water. At that point, we lost our primary electrical feed. Once that happened, we only had 2.4 megawatts of peaking generator in addition to our smaller building sized emergency generators. This required managing our load, which is normally 8 megawatts, down to 2.4 megawatts.
The peaking generators did not have transfer switching and the local utility company had to manually switch us off their substation in order to not back feed their lines. When the utility company was able to reestablish their primary feed the entire process had to be reversed. This happened several times. The peaking generators consumed tremendous amounts of fuel, and all of the oil companies and delivery trucks were cut off from us north of the Tar River.
We had a million gallons in our tanks, less than a hundred yards away, but no way to transport fuel except by stringing a hose across railroad tracks (luckily the trains were unable to run). Our primary steam plant emergency generator had a 10,000-gallon "day" tank, but we did not have a good method of automatically transferring fuel to it from the bulk storage tanks.
What we were learning was that we had good short-to medium-term emergency capabilities, but lacked the ability to sustain emergency operation of our backup systems for an extended period.
We also found our load management plan to be out of sync with the needs of our phone and computer networks. Our plan had been to bypass all unoccupied and non-essential buildings. While we had emergency generators on our main telephone switch, the fiber networks were a distributed node system that required a small amount of electrical power in a large number of places on campus. This was not obvious at first because the battery backups for the nodes kept everything going for quite a while.
Once we figured out why the systems started to fail, we had to develop a load management plan that energized almost the entire campus electrical system, but shed all unnecessary load in each building. We assigned it to one of our engineers and told him when we needed a solution. He was the vice president of load management and responsible for all decisions, priorities, and tradeoffs.
We found we had to modify our normal management approach. We found the sheer magnitude and number of real-time problems facing us too great to handle using our normal approach, we had to distribute our problem solving horizontally. We found that someone needed to focus on keeping the operation going while most were trying to solve the remaining problems. There are normally discussions, approvals, etc. in our operation when we make critical decisions, but there simply wasn't enough time for this luxury.
The next thing to go was the potable water supply as the city's water treatment plant was slowly overcome by the rising floodwaters. That caused us to switch to bottled water and the dining facilities had to modify their food handling practices-the steam plant that supplied them with hot water became ever more critical.
We found that priorities changed hourly as we felt the impact of external forces. We started speaking of the "plan of the hour," not the plan of the day.
As time moved on, the Campus Emergency Response Team grew in number as the membership proliferated. We basically outgrew our meeting room twice. We found ourselves doing a lot of on-the-spot problem solving. As one would guess, less and less of the output of the group was applicable to facilities from an operations standpoint. This was not at all inappropriate, because other issues gained priority as we solved the facility problems and the outlook of reopening as scheduled looked brighter.
This did result in "the meeting after the meeting" to develop an operating plan for the next 12 to 24 hours based on decisions made in the large meeting. This group consisted of the executive director of facilities, director of housing, director of dining services, director of environmental health and safety, the assistant chief of university police, and me. The group met after the large meeting and developed an operating plan. The first of these plans was literally written on the back of a dining table information folder.
The plagues didn't seem to stop. The water that was no longer potable ceased to flow at all. Probably one of the smartest decisions made early in the week was to order and stage 60 portable toilets. By the time we needed them, they would have been impossible to get.
One of the final plagues was the failure of the local utility's sewage lift station, near the university. At the crest stage of the Tar River flooding, it was literally trying to pump the river. This resulted in one of the worst cleanup sites on campus.
There were many heroes throughout the university, some of whom were dealing with the 2,000 students that no longer had a habitable apartment, or books, or computers. Yes, Hurricane Floyd did "eat their homework." Some were dealing with relief efforts for staff (the hundreds of university employees who lost everything). At one point in all this, the Methodist Retirement Village, near the university, was overtaken by the flood waters; university buses transported the elderly residents over a hundred miles to another center. We had employees come to work from emergency shelters, the only spare clothes they had were our uniforms. One employee, after rescuing his dog from his flooded apartment, kept Zantar at the shop and slept in his truck so he could keep working.
What else did we learn?
Have a Plan
I know that sounds obvious, but I say this after already telling you, we did not have a specific plan for such an unprecedented event. We did have a Hurricane Plan, Winter Storm Plan, and a Y2K Disaster Recover Plan. It turned out that our Y2K "Doomsday Plan" readiness exercises and "what if" scenarios provided some of our best guidance. The Y2K Plan dealt with the effects of wide area disruption of utilities, supply lines, and communications. But be sure to look at how long you can sustain your plan.
Don't Fall in Love With Your Plan
Use your plans as a guide-modify your plan to fit the circumstances and remain flexible. We found that often the "Plan for the Day" changed hourly as external forces and events came into play. There is no substitute for a focused, highly motivated management team when it comes to real-time problem solving.
Communicate the Plan
Make sure the whole team is running the same play from the same playbook, as they say in football.
Radios
Our VHF radios are our primary communications link for operations. Make sure repeaters are on emergency power along with chargers for batteries. Have lots of spare batteries.
Phones
Keep the phone system up. While it's not as important to facilities, it is for almost everyone else.
Network
The power of the Internet and web-based information really came into its own during this event; it was the primary medium for announcements and information for students, staff, and especially parents.
Cell Phones
Cell phones were the secondary source of internal facilities communications, but a primary source of external facilities communications.
Don't Change Anything that Doesn't Need to be Changed
This may sound like a contradiction to my earlier statement of "Don't Fall in Love with the Plan," but it isn't. Use as many normal, pre-established systems as possible. We use our normal protocol unless we lose our repeater, then we have a pre-established plan to go to a simplex approach. We continue to operate our Facilities Service Center and answer our phones. However, our published instructions to the campus indicate that we should only be called for emergency items.
Lists: Make Them, Update Them, Keep Them Handy
There is a reluctance to make lists that seem to be obvious. Our emergency plans include lists of generators, what is supported by the generators, fuel capacity, and run time between fueling. While this is an obvious list, others were not, and some had to be developed. We found we needed a list of the air conditioning systems that would work without domestic water being available. We didn't have such a list; we do now.
Yes, we have people who know all those things in their head, but remember that a great many key members of your corporate memory may be missing. Things that are easy to remember during normal times may not be that easy when your people have been working 16 to 18 hours per day at 150 percent effort, for several days.
FEMA & Insurance
This is one area we had just developed procedures for handling. That turned out to be invaluable. Everyone needs to understand the data collections and forms before the emergency begins. There isn't time to learn on the job! You have to start collecting information and documenting as things occur or it is too late. You have to have someone on your staff that is a "FEMA expert."
There are so many people to thank for making our recovery possible. Our business partners were unbelievably generous. Sodexho Marriott Services sent people, money for food, and relief supplies for our employees. Johnson Controls sent a technician and a truck full of parts for two weeks. Aramark kept our workforce well fed, and I don't just mean peanut butter sandwiches. Many other vendors donated equipment and workers. Other universities called and offered to help with anything we needed. If they had it, it would be on the way. North Carolina State University, UNC/Chapel Hill, UNC/Greensboro, and others were all a part of our speedy recovery.
Without a doubt, Hurricane Floyd took a toll on our organization. We operated in what felt like a war zone, for 20 hours a day, on adrenaline. We started to transition back to regular operation while continuing certain emergency mode activities. What was really hard was redefining and reinforcing-both for our management team and our workforce-what was important: that mowing grass was again important, that cleaning mirrors was important.
Getting back to normal took much longer and required much more effort than I expected. About six months after the event operations seemed to have returned almost to normal, but perhaps things will never be the way they were before Floyd.