By Paul Kirvan
Despite all the precautions you can take to provide uninterrupted power to your data center, situations can occur that threaten your center's continued operations. Emergency power strategies should be an important part of your disaster recovery (DR) planning. Without emergency power systems and strategies to protect your power supply, your data center is significantly at risk.
Assuming a power system is reliable, we can probably assume that it will also be available. While no power system is infallible, we can improve the chances of its continued operation by making appropriate safeguards. The goal is to minimize the potential for component failure. One useful way to do this is to identify and address potential single points of failure. This can be achieved through power audits and assessments. If the failure points can be identified and the threats mitigated or eliminated, we can increase power availability and improve equipment reliability.
DATA CENTER EMERGENCY POWER STRATEGIES IN DISASTER
Table of contents
The impact of power disruptions on your data center
Power disruptions or failures can cause the following to occur in a data center:
- Partial or complete center shutdown, or below-standard operation
- Unacceptable performance of data center equipment
- Activation of electrical protective relays or emergency operation of backup electrical systems
- De-energizing of electric circuits or equipment
Electric power companies are major partners in data center operations. Close cooperation with carriers and regular reviews of power quality will prove beneficial. As power quality can vary greatly by provider, it is essential to invest in equipment that removes or minimizes power anomalies such as voltage/frequency fluctuations, sags, spikes, surges, brownouts or blackouts. This includes power conditioners, line filters, surge suppressors, grounding/bonding and many other devices. If it's possible to obtain primary commercial power from two different power grids, and have that power routed to your data center via diverse paths, you can improve your chances of recovering in a power outage. However, the cost to engineer and construct such a diverse power infrastructure can be prohibitive.
Emergency power system components
In a medium to large data center, emergency power systems typically include a centralized uninterruptible power system (UPS) providing continuous power in the event commercial power becomes unavailable. When a commercial power interruption is sensed by the uninterruptible power system, it automatically sends a signal to power up the generator while at the same time maintaining the data center electrical load on a string of batteries. Once the generator has come online, it assumes the electrical load. Be sure the generator has enough fuel to last at least 48 hours fully loaded, and that multiple fuel suppliers are available. As long as the tanks are refueled, the generator can run indefinitely. The generator should provide emergency power to handle computer loads and the data center's air conditioning, plus telecom closets, emergency lights and other loads as needed.
Sizing the emergency power system is critical, so that the initial configuration should be able to handle the anticipated loads. Since many uninterruptible power systems are modular, the backup power environment can be expanded via additional uninterruptible power system modules and batteries. Be sure to size the generator to handle initial and potentially increased requirements.
Besides providing a smooth transition to emergency power during outages, the uninterruptible power system conditions power to remove spikes and sags that can damage computer components. Power coming from commercial sources is not clean enough to provide the consistent power required by sophisticated electronics. Both UPS and devices mentioned earlier (e.g., conditioners, filters) can mitigate this situation.
To ensure that the generator will work when needed, perform regular run tests, especially with a medium to full electrical load. Consider installing emergency power systems equipped with load banks (special circuitry designed to consume energy) capable of providing loads equaling 100% of the generator capacity. This allows full testing without impacting data center operations.
Dealing with data center power emergencies
When something interrupts power system operation, emergency procedures are necessary for a speedy resolution of power problems while minimizing the impact on critical data center systems. Such procedures should list, step-by-step, actions to be taken for a given type of emergency. If these procedures are not available, be sure to have access to trained maintenance personnel so a response can be organized. Assuming that your data center maintenance is contracted, you may have minimal on-site staff available to address emergencies. It's possible that if on-site employees are not familiar with power system operation, equipment manufacturers will need to supply this knowledge. The result is that when a power system emergency does occur, even if emergency procedures are in place, there may be nobody familiar enough with the primary and backup power systems to properly implement them.
Adequate and regularly updated power system documentation is essential. Data centers are dynamic environments, and as critical systems and additional infrastructure to support them are added this documentation must be maintained. Be sure to locate your primary and backup power systems in secure areas to prevent unauthorized access.
Perhaps the most important strategy for protecting power systems is regular maintenance. This means scheduling tests of primary and backup power systems, regular inspections, and following manufacturer recommendations for maintenance and support. Another key aspect of maintenance is the need for benchmarking. During maintenance, various tests are performed. The results of such tests are the most meaningful when they are tracked over time, rather than simply counted as "pass/fail."
Commissioning of power systems within the data center is another protective strategy. True commissioning is different from simple component startup, which only tests the component in question, usually with predefined vendor procedures. Ideally, test systems end-to-end across the data center to make sure all components work together properly.
Today's sophisticated data centers handle mission-critical operations and processes, and it's not feasible to shut them down -- even for a short duration. This means that power needs to be available continuously. A properly designed and regularly tested emergency power system will ensure that critical data center operations are protected.
You must have Adobe Flash Player 7 or above to view this content.See http://www.adobe.com/products/flashplayer to download now.
Download for later:
Podcast: How server virtualization benefits disaster recovery
• Internet Explorer: Right Click > Save Target As
• Firefox: Right Click > Save Link As
About this author: Paul F. Kirvan, FBCI, CBCP, CISSP, has more than 20 years experience in business continuity management as a consultant, author and educator. He is also secretary of the Business Continuity Institute USA Chapter.
This was first published in May 2009