Nothing guarantees end-to-end power system assurance for critical power loads. However, criteria and good practice can raise your chances of success.
What is end-to-end power system assurance? We define it as a comprehensive, appropriate, logical electrical system having related disciplines providing: simultaneous maintenance and operation without incident; transparent load operations; continuous operation (24 hours a day, seven days a week); and end-to-end (from power source to critical loads) reliable power.
Although these criteria are ideal, no system ever achieves a 7/24/forever basis. To reach this goal, it would take high-quality utility service, UPS system(s), a standby power plant, and a transfer switching arrangement, all working together in harmony forever. Yet we can come close to a system solution by blending equipment specifications with efficient and safe operation/maintenance.
The nature of failures. Before determining how to design a reliable electrical system, you must understand where and how system failures are likely to occur. Similarly, you must understand how each of the components providing dependable power can fail, and how each operates as a part of a much larger system. It's interesting to note that you can trace more than 64% of all failures to human error. The man-machine interface on today's equipment is much more complex, which may lead to system failures, especially when the system changes its mode of operation or recovers from a bypass position.
A perfect solution for a high-reliability power system is hard to find due to cost, functionality, operations, and maintenance. As you drive reliability higher and higher, cost escalates rapidly.
Providing assured power system components. The most effective means of analyzing a complex power system assurance approach is to break the system down into small pieces and examine each carefully. The areas of the electrical system you should always consider when seeking a high-reliability solution are:
• Incoming utility services;
• Engine-generator plant configuration and specification;
• Load feeders and selection;
• Normal-emergency transfer schemes;
• UPS system topography and transfer modes;
• UPS/critical power distribution;
• Fire detection systems;
• Building monitoring and automation systems and approach;
• Supporting systems such as clocks and paging/communications;
• Lighting;
• Mechanical system support; and
• The man-machine interface.
The trick is to balance the technical merits and risks of each of these with the cost, schedule, and associated maintenance, to provide the best solution.
The elegance and effectiveness of your design comes from the blend of these systems. Because of design and construction cost expectations, you should first focus on all single points of failure. Then, should the budget and schedule permit, address the more esoteric modes of failure.
One of the most powerful cost-control tools available is the ability to defer the purchase of equipment, based on the type of electrical distribution system design you select. With dual-path power systems, you can save money on your project by not purchasing UPS/critical power and generator systems. If you need them later, you can transparently and safely add the systems to the existing systems. But this alternative won't fit in many applications.
Offering a concurrent, maintainable, and operable approach. Before you undertake a project, decide the type of electrical distribution system you wish to use. You can essentially choose one of two options: concurrently maintainable/operable or fault tolerant.
An assured power system must provide a dual-powered path to your critical loads. Centralized mechanical-electrical-plumbing (M-E-P) systems, such as UPSs, engine gen-sets, and chillers, are redundant. While we typically consider the systems to be n+1, this is misleading because several system topologies offer unparalleled redundancy.
These systems offer some degree of fault-tolerance during normal operations; however, the design doesn't always offer redundancy under some maintenance operations.
The first type of assured power system (concurrently maintainable and operable system) allows you to operate and maintain the M-E-P infrastructure concurrently, while connected to the loads. This occurs in a completely transparent and risk-free manner to the loads. This means the provision for periodic and routine maintenance incorporates into the system topology, which respects all switching modes and a minimum of a single failure in a system. This type of system may not provide failure recovery or system redundancy during some maintenance operations. The uptime goal for a concurrently maintainable and operable system would be 5th Sigma, providing 99.999% uptime.
The objective of a fault-tolerant system is to use all the design features of a concurrently maintainable and operable system, while adding benefits, such as system redundancy during switching and maintenance operations as well as an ability to automatically switch the loads.
Two visible features to this system include: using static transfer switches for critical load source arbitration, and retaining redundancy in the main M-E-P systems when units and systems go off-line for maintenance. This means you can maintain and operate the fault tolerant system when undergoing all maintenance operations. The uptime goal for a fault tolerant system is in the 6th Sigma or 99.9999% uptime.
Key points to remember. Here are some design and operational points to remember for your next project:
• Eliminate all single points of failure.
• Determine if you will have a fault-tolerant or concurrently maintained and operated system.
• Understand that the higher upstream a failure occurs in the system, the more widespread it will be. The lower downstream the failure occurs (or the closer to the load the failure is) the more acutely it would be felt by the load.
• Specify equipment that solves known sources of component failures like generator starting systems and control power systems for breaker operation.
• Make sure all changes in the system's mode of operation (as defined by going from a steady-state condition to an alternate operating posture) are clear to an operator, safe to perform, and doable under duress.
• Provide primary and alternate feeds to all critical, cooling, lighting, and supporting systems.
• Provide primary and alternate UPS feeds to all critical loads with some form of sub-cycle load transfer. You may accomplish this transfer by a system-provided automatic sub-cycle transfer switch or by the power management system within the load itself.
• Examine the paralleled operation of all large-scale, enterprise-level systems such as generators, utility services, and UPS systems.
• Design the ampacity level of the electrical feeders to the redundancy of other systems. For example, if the HVAC system requires four out of every six air handlers must operate at all times to support the critical load, the electrical system design needs to account for this requirement during normal, maintenance, and failure modes of operation.
• Provide system switching operations that will not result in an overload condition.
• Provide two UPS feeds to all critical loads with automatic load arbitration by either an external ATS or by a load's internal power management system.
• Assure that the supporting cooling, fuel, and fire suppression systems are adequately sized and configured to support the electrical system.
• Train staff for proper and safe system operation.
• Make sure you have complete operation and maintenance documentation for all equipment, and that the documentation is at the job site before final testing and acceptance.
Following these design and operational points can bring you closer to achieving an uptime goal in the 6th Sigma range. Although a perfect solution is hard to find, understanding how a system fails and providing assured power system components can help you to raise your chances of success.
Mazzetti is a Principal with Mazzetti & Associates in San Francisco.
Sidebar: Failure Characteristics
You'll find that many assured power system designs seek to prevent the "knockout punch" of failures. However, you may not hear about the many types of failures actually occurring. These failures have one or more of the following characteristics:
• More than 64% relate to human error in judgement, or failure to follow published guidelines, procedures, or directions.
• Failures are often due to a deteriorating condition in either an OEM component or a related system that cascades to the failure. The root cause may not always be in the equipment. Therefore, the MEP monitoring system provides critical status information on an ongoing basis.
• Designs typically don't address common failures often experienced in the equipment specified. For example, the main source of failure in generators is a failure to start. So, a major focus of the specification should deal with the starting batteries and motors as well as the fuel filters.
• Designs often do not include all maintenance modes or the ability to recover from maintenance operations.