NFPA Journal®, September/October 2009

By Lisa Nadile

In the two years since 32 students and faculty were killed at Virginia Tech in Blacksburg, Virginia, many universities and colleges across the country have taken steps to beef up their emergency communications systems.

But as experts are quick to point out, merely installing the latest gear isn’t enough; you need to know how it works as part of a larger emergency response plan. You need to know what the right message is for any given situation, and you need to know the best way to communicate that message.

When the Virginia Tech shootings took place on April 16, 2007, the university was already preparing to upgrade some of its emergency notification technology, including its text messaging system. The shootings made the school realize that, despite all the technology, the emergency notification plan itself needed work. Campus officials needed a better way to synchronize the emergency tools they already had to get the right message to as many people as they could as quickly as possible.

“We’ve always had notification protocols because we live in a part of the country where you sometimes have to close because of bad weather,” says Larry Hincker, assistant vice-president of university relations and self-described custodian of Virginia Tech’s emergency notification system. “So we had some vehicles before that we would use, but...none of it had ever really been coordinated and centralized.”

Many institutions are recognizing the need for that kind of coordination, and are taking a close look at their emergency protocols. For everyone, the goal is to make campuses safer and more manageable in the event of an emergency.

Hard lessons
On the day of the shootings, Virginia Tech had a number of mass notification systems, including audible sirens, a telephone hotline, a dedicated emergency Web page that could be updated quickly, and the university email system, which had 36,000 accounts and could deliver 10,000 emails per minute. In the aftermath of April 16, Virginia’s governor convened a review panel to analyze the incident and the university’s response, and the panel’s findings include an analysis of the school’s emergency response plan .

There were several deficiencies, the panel stated. A shooting scenario was not included in the plan, and the police did not place high enough in the decision-making hierarchy. The panel also stated that the protocol for sending an emergency message was cumbersome and untimely, and noted that university administrators waited hours before sending an emergency communication. In this situation, the effectiveness of a quicker shelter-in-place alert is purely speculative, but administrators say a hard lesson was learned. “We learned there is an inverse relationship” between the urgency of the incident and the timing of the message, says Michael Mulhare, P.E., Virginia Tech’s director of Emergency Management. “The more important it is to get the message out, the shorter the time frame in which to send it.”

Virginia Tech undertook a redesign of its mass notification system. First, it reassessed the threats facing the campus, evaluated its mass notification technologies in terms of those threats, and took a hard look at its plans for using the system. It then began reworking the plans using two concepts: the system should be appropriate for all hazards, even the rarest, and it should be redundant, sending notifications through as many channels as possible to overlap the communication habits of the campus population.

The new system, named VT Alerts, allows students and staff to subscribe to the information delivery methods of their choice, be it a text message to their phones, an email, or an announcement that flashes on their computer screens using a downloaded piece of software. They can also choose multiple methods, and prioritize those choices. Tied together into one system with these individual alerting modes are telephone switchboard announcements, traditional sirens and loudspeakers, posts to a special announcement area of Virginia Tech’s Internet home page, and a weather and emergency hotline.

The school has also installed 500 digital message boards, at $1,800 apiece, in all classrooms and in high-traffic areas throughout the campus. According to Hincker, the signs are popular because they offer students and staff a sense of comfort. Hincker is one of about 30 university staffers trained on the system, and he can access a website and create emergency messages for the signs using customizable templates.

While Virginia Tech devised VT Alerts on its own, the process for creating a detailed emergency plan is contained in NFPA 1600, Disaster/Emergency Management and and Business Continuity Programs, and in FEMA materials, notably in its publication Building a Disaster-Resistant University. The FEMA models are also used by Georgetown University in Washington, D.C., to structure its response plan and emergency response team, says Whit Chaiyabhat, director of Emergency Management and Operational Continuity for the university.

A member of NFPA’s Emergency Communications Systems Committee, Chaiyabhat worked on the new chapter on emergency communications in the new NFPA 72®, National Fire Alarm and Signaling Code®, which also requires a risk analysis and emergency plan. “Looking at the various technologies that we have available to us, it’s necessary to have some sort of standard available towards the technical build-out, design, installation, and maintenance of these systems,” Chaiyabhat says.

Like Virginia Tech, Georgetown has completed a detailed risk assessment, classifying each hazard according to one of four threat levels. Each level has designated plans and responses.

“We looked at each hazard or event and, based on the various technologies and systems we have available currently and the message delivery modes, we decided what we would say with each system,” says Chaiyabhat.

One unique aspect of Georgetown’s current notification system is the Campus Alert System, a network of steam whistles throughout the campus that the university has reused as a sort of loudspeaker. “It sounds just like a train whistle, and it means only one thing: take shelter inside immediately,” Chaiyabhat says.

Georgetown created notification procedures, called incident management flowcharts, and created a group of key people who rotate as the senior administrator on call. Departments that receive an initial emergency call notify the administrator following existing protocols, and the administrator serves as a one-person emergency operations center.
“That individual basically makes the decisions with that initial call,” Chaiyabhat says.

Crafting the message
Like Virginia Tech, Georgetown disseminates information using a multiple-channel approach, broadcasting emails and voicemails, using a weather hotline and local broadcast media, and updating its emergency website (http://safety.georgetown.edu). The university also has a subscription-based text messaging service called HOYAlert, named for the Hoyas, the nickname of the university’s sports teams. It also has a network of trained staff and students that can be reached using radio or email and are charged with communicating with others. “We use these folks to create redundancies in our alerting capability,” says Chaiyabhat.

According to Virginia Tech’s Hincker, communicating with students and staff using the channels they use is key to ensuring emergency messages are seen or heard by everybody on campus. For students, these channels include the social media networks Facebook.com and Twitter.com. The information Virginia Tech posts to its Facebook and Twitter Web pages is not only received but shared by students and staff who might not be linked with the university’s accounts, he says.

According to Wayne Moore, principal with Hughes Associates, a fire protection consulting firm, an emergency communication message should state the hazard and its location, and specify what action to take. It should also make clear who is sending the warning. This calls for concise statements, since most text messaging systems limit the number of characters that can be sent per message. For example, text-to-phone messages can only be 160 characters long, including spaces.

The prominence of emergency communications systems using Internet technology raises other issues, chief among them that a college’s information technology (IT) department is now a critical part of the school’s mass notification plan. The department will have to ensure that its networking and telecommunication resources will support an emergency. As Hincker puts it, “You’d better take your IT leader to lunch.”

With so much depending on IT systems and the Web, some experts, including Chaiyabhat, favor more discussion around performance standards for those systems. Companies including IBM, Microsoft, Verizon, and Sun Microsystems are working with government agencies such as FEMA and the U.S. Department of Defense to develop a data interchange software standard dubbed the Common Alerting Protocol (CAP). Adherence to a communication standard such as CAP would make these new technologies more plug-and-play, rather than requiring customized networking solutions that may downgrade performance and be costly to implement, says Chaiyabhat. (For more information on CAP, visit www.oasis-open.org.)

In reworking its emergency plans, Virginia Tech is striving to help the campus recapture a sense of wellbeing. And the plan has already been tested. In January, when a graduate student was murdered in a campus café, the university used VT Alerts to notify the campus of the incident and to send the message that the assailant was in custody.

“We thought that certainly there was a need to notify the university population,” says Mulhare. “The information would start percolating on its own, and we wanted to make sure we got the correct information out.”

SIDEBAR
Wanted: Funding for messaging research

The content and characteristics of messages transmitted by emergency communications systems are critical for their effective operation, and facility operators, including college campuses, are raising important questions about current messaging strategies. These questions involve the level of consideration that should be given to characteristics such as the threat or hazard, the stage of the event, the sources and recipients of the communications, recommended message content and format, and the methods of message delivery.

A new research project addressing these questions is currently being proposed by the Fire Protection Research Foundation FPRF). The project, “In-Building Communications Strategies,” is intended to develop guidance information to help implement effective communications strategies for a variety of hazards and threats. Project deliverables will focus on addressing indoor notification requirements, which may or may not be part of area-wide notification initiatives.

The 2010 edition of NFPA 72®, National Fire Alarm and Signaling Code®, includes a new chapter with important expanded requirements on emergency communications systems. The work of the NFPA 72 Technical Committees has focused on providing “menus” that permit the development of communications strategies based on differing levels of risk for a variety of hazards and threats. Earlier studies relate to this topic, but do not fully consider important influencing factors, including recent major advancements in mass communication technology; the state-of-the-art systems now available for emergency communications; shifting demands to address emergencies other than fire, such as security events and natural catastrophes; and updated studies on human behavior during emergencies.

Funding for this project is pending. Anyone interested in supporting the project should contact FPRF Program Director Casey Grant at cgrant@nfpa.org.