|
The connection to this fuse failed before a survey could be carried out. In a critical application, that could have led to unwanted downtime. |
Colin Pearson considers how to maximise the reliablility of systems serving critical environments
Building operators need to get more and more out of their assets. This means that they must use their plant and equipment as much as possible for as long as possible — or, as some say, ‘sweat their assets’. They must often get as much as 99.999% (sometimes known as ‘five nines’) reliability from their plant. They may even require reliability up to ‘nine nines’, or less than 30 ms a year downtime, for critical business-support systems such as electrical supply.
These requirements bring a new set of values to many facilities engineers. Traditional maintenance with annual shutdowns is no longer possible. Systems need to be more robust, and ‘single points of failure’ where the electrical supply depends on, for instance, one set of busbars, must be designed out. Complete dual systems may be required to ensure reliability. Downtime is just not acceptable. The old adage that the quickest way to learn is from your mistakes is not good enough. You must learn from the mistakes of others. This means networking with others in the same field to find out what they are doing and why.
The Critical Environments Network, set up by the Building Services Research & Information Association, aims to provide the information that people need to build, operate and maintain critical environment-support systems.
Critical environments include data-processing facilities, call centres, e-commerce businesses, telecommunications switching and network hubs, pharmaceutical plants, transportation centres and museums. Engineers who design and operate services for financial institutions are already joining this network, but almost every business has critical environments. Solutions developed for the pharmaceutical industry, where a few seconds downtime can jeopardise a 5-year research programme, may be equally useful in data centres.
Design
The common factor in all critical environments is their dependency on a reliable infrastructure. Designers need to know the criticality of each service for each project and how to design for 100% reliability. Single points of failure must be designed out. The design of critical systems is a specialist skill, and a designer with suitable experience should be found.
Build
The quality of materials, plant, equipment and installation work is a key element in achieving reliability. Off-site manufacture, accurate delivery scheduling and a clear space for installers to work on site are also important elements. Motivation to produce a quality installation is essential. Finally, commissioning must not be forgotten or left to the last minute; it must be planned and managed.
Operate
The operation of a critical environment cannot be left to chance. Success depends on the three Ps — plant, process and people.
The right plant should already have been installed, but the process of operation needs to be developed and conveyed to the operators. Good handover of information is essential. Finding the people with the right skills to operate and maintain critical systems is increasingly difficult. However, specialist courses are available for most systems, and staff should attend these and refresher courses to ensure skills are kept up to date. Courses can often be arranged cost-effectively on site if there are more than three people to be trained.
Maintain
Planned maintenance on the basis of predicted life is no longer good enough because it leads to unnecessary replacement. However, downtime through component failure is unacceptable. Loss of running time on a chiller system, for instance, can mean overheating, computer failure and loss of business. Most of the plant cannot be run to failure. The ideal solution is to run the plant until just before it fails!
In these critical environments, conditioning monitoring is the only way forward. More and more building operators are finding that they can run their plant for longer if they are monitoring its critical failure modes.
But what are the critical failure modes?
What will fail first?
How will it fail?
What will happen when it fails?
And how can it be repaired?
These questions must be answered before a condition monitoring programme can be started.
Condition monitoring is not an end in itself. It can only reduce the bottom line of the maintenance budget if it is part of a condition-based maintenance regime. It is as much about management as maintenance, and many who use it now refer to their role as conditioning management.
BSRIA has long seen the need for a more intelligent approach to maintenance and published guides to ‘Business focused maintenance’ and ‘Condition based maintenance’. To pursue this further and offer a dedicated service, BSRIA has now set up a condition-monitoring department. Condition-monitoring services include thermal imaging and acoustic emission, but the key element is helping customers to set up condition-management regimes in their facilities.
Colin Pearson is head of conditioning monitoring at the Building Services Research & Information Association, Old Bracknell Lane West, Bracknell, Berks RG12 7AH.
[colin.pearson@bsria.co.uk]
