Design masterclass 5 Statistics
Avoiding overdesign
Masterclass Professor Doug King
This month we look at how we can use statistics to determine suitable design parameters for efficient systems
I
f we want building systems to be efficient in their use of materials and operational energy we must avoid over-designing them. There is constant pressure on the building services engineer to over-
design, which must be countered at every turn. This comes in part from a culture of risk avoidance in the industry and in part from a client or developer being unclear on how the building is to be used with the engineer responding by designing for all eventualities. However, operational flexibility beyond a provision that can be reasonably expected always leads to over-design and inefficiency. As intelligent engineers we are able to exercise
judgment based on experience or precedent. However, when presented with a new circumstance, which all projects do to some extent, we need to be extremely careful to avoid trying to eliminate all risk. It is not possible to completely avoid the risk that, under extreme circumstances, the systems will not be able to perform to the desired level. What we need is a way to establish an acceptable level of risk given the consequences of failure of the system to perform as required. The consequence of occasionally being unable to
deliver the full design flow of hot water is very different to the consequence of a major international airport being unable to function because of a predictable weather occurrence, and so we must evaluate the risk and service levels accordingly. In order to solve these questions we need a basic understanding of statistics, but the reward can be considerable savings in terms
www.cibsejournal.com February 2011 CIBSE Journal 51
of initial investment and running costs with a small, acceptable level of risk of non-performance. These types of statistical analyses are applied widely
in the building services field, although we often don’t realise it. The ’demand units’ method for sizing piped services is based on a statistical analysis of the simultaneous occurrence of a number of independent events. If we understand the fundamental approach we can estimate diversity factors for a wide range of design problems. If, for example, we were asked to design an office
for a firm of estate agents, should we design the air conditioning (assuming that we have not achieved our design by passive means) for 100% occupancy or some lesser number based on the knowledge that estate agents are frequently out of the office. If we discuss the working patterns with the client
we might discover that on average each viewing lasts for half an hour and that visits are entirely randomly distributed between the agents and throughout the day. So we can be fairly certain that the office will not be fully occupied, but the derivation of occupancy is not
>
Even in an entirely conventional office it is unlikely that there will be continuous 100% occupancy due to absences, breaks and meetings. How, therefore, should we determine the optimum design level for building services without over-designing?
Doug King
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72