Comfort and productivity - Guide lines

Building Services Journal invited 30 leading industry practitioners to a one-day workshop. The resulting discussion was recorded and is partly reproduced here to aid readers' understanding of both the content and context of the Guide A proposals.

While the event was not a formal part of the revision process, the comments made by the delegates will inform the final version of Guide A. Readers should be aware that specific criticisms may become redundant.

Section A1: Environmental criteria for design

This rewritten section of Guide A retains the same basic structure as the 1986 edition, covering the thermal environment, humidity, outdoor air supply rates, visual environment, acoustics and noise, and vibration. Two new parts have been added to Section A1, one on the electromagnetic and electrostatic environment, and a major section covering indoor air quality.

The first major addition to the 1986 edition is Table 1, which lists recommended comfort criteria (environmental temperature, outdoor supply rates, filtration, maintained illuminance and noise rating) for air conditioned buildings. The new table effectively subsumes several tables in the 1986 edition and part of it is reproduced here.

For the first time the Guide acknowledges different comfort criteria for non-air conditioned buildings, stating that, "Higher temperatures may be acceptable...if for sedentary areas such as offices, an inside dry resultant temperature of 25°C is not exceeded for more than 5% of the annual occupied period."

Guidance on relative humidity has also been altered. An upper limit of 70% rh is based on the point at which microbial growth and proliferation of house dust mites is likely to occur. The lower limit of 40% rh is the threshold for electrostatic discharges and airborne irritants.

Although the wording of this section permits relative humidities of 30% for short periods (with the proviso that precautions are taken to limit the generation of dust and static discharges), the 40% lower limit still came under fire from delegates concerned at the potential for mis-interpretation.

Delegates warned the CIBSE not to express guidance in a way that might mislead designers. "Junior members of the profession might be tempted to treat the figures as rigid limits and thereby install humidification unnecessarily," warned Brian Mark. While section author Paul Appleby stressed that the 70% and 40% rh values should not be regarded as design levels, he conceded the point and proposed to alter the wording.

Indoor air quality

The new section on indoor air quality generated the greatest interest among the workshop delegates. This section of the Guide is in recognition of the thousands of airborne contaminants now known to be a feature of buildings.

The Guide should say something on smoking, we can’t just ignore it

Mike Holmes

Many of the substances listed in the draft Guide A are considered by the World Health Organisation (WHO) to either cause discomfort or be harmful to health. "We decided to include the recently revised air quality guidelines produced by the WHO," explained Appleby. "These provide threshold levels for specific pollutants based upon potential ambient exposure.

"While the guidelines are essential target levels for local authorities to achieve, rather than for the use of building designers, they set a health standard to aim for," he added.

The CIBSE's approach is similar to that adopted by the ASHRAE in its revision of Standard 62.1989. However, the Guide differs markedly from the standard in its advice governing smoking environments.

The ASHRAE has abandoned attempts to set different ventilation rates for smoking and non-smoking environments, under pressure from the smoking lobby. The draft CIBSE Guide is more brazen in setting ventilation rates for smoking levels; non-smoking environments are maintained at 8 litres/s/person, while continuous heavy smoking requires 43 litres/s/person.

The section suggests that if the pollutant cannot be eliminated at source, attempts should be made to either reduce emissions, dilute the concentrations by increased ventilation or segregate occupants from the origin. For pollutants emitted at both steady and non-steady rates, the Guide now provides equations to arrive at an appropriate level of ventilation.

Several delegates expressed doubt about any mention of smoking in the Guide, given the ASHRAE's ructions over Standard 62-1989. "I think that it is very dangerous for the CIBSE to even mention smoking," said CIBSE senior vice president David Wood.

Consultant Brian Warwicker said that the Guide should only mention 8 litres/s/person for non-smoking environments "and let designers make up their own minds what the level should be for smoking environments."

Ronald Kirkwood thought it a "good idea" that the CIBSE should avoid taking any responsibility for saying that a building is safe by doubling the percentage of fresh air. Ove Arup's Chris Twinn suggested an amendment. "Perhaps we should say that the rates are suggested values and do not necessarily indicate comfortable and healthy conditions," he said.

In defence of Table 1.10, Arup's Mike Holmes reminded delegates that many CIBSE engineers design pubs, clubs and restaurants, where smoking is allowed. "The Guide should say something on smoking," said Holmes, "we can't just ignore it."

The visual environment

While the section on the visual environment is very much in line with the existing Guide, a new section has been added giving criteria for design using daylight. This covers issues such as light distribution, permissible average and minimum daylight factors and notes on the directional effects of daylight, particularly where display screen equipment is used.

Without repeating everything that is said in Lighting Guide 3 or Applications Manual AM² Window Design, the revamped Guide A does stress that designers must take account of daylight direction.

Acoustics

If we can’t be accurate on site, perhaps we need to match system performance closer to how buildings really perform

Gay Lawrence-Race

The section on acoustics and noise has been revised to include what are called room criteria. This stems from developments in North America. Noise reduction (NR) curves were never adopted in the USA, and indeed the ASHRAE no longer recommends noise criterion (NC) curves as a design criterion. Instead the US uses room criterion (RC) curves, and the CIBSE has decided to include the RC method to give designers more options.

RC curves are based on actual measurements in air conditioned buildings judged by the occupants to have good acoustical environments. The result is a set of parallel lines falling from low to high frequencies at -5 dB/octave. Noise spectra which follow these slopes are said to be acoustically neutral2.

For designers, the RC curves tend to be more sensitive to low frequency sounds, such as noise from chiller plant. As many American air conditioning manufacturers quote RC curves in their specifications, their inclusion in Guide A should make designers' lives that bit easier.

Section A2: External design data

For external design data read weather and solar data. Last revised in 1982, this section has received a major rewrite, and indeed will be the subject of a separate publication, Guide J, to be published shortly³. Hence the weather and solar data given in Guide A is a condensed version of that in Guide J.

The big change since 1982 is undoubtedly the use of computerised dynamic simulation which has advanced to the extent that the CIBSE will be providing weather and solar data specifically for simulation, with data for manual calculation provided as backup.

"This section has been written to enable designers to sit down with clients and discuss the options, rather than rely on prescriptive temperatures of 21-22°C," explained principal section author Geoff Levermore.

"Simulation run-times dropped rapidly during the revision of Guide J, and as a result there will be more client-based risk analysis carried out for which a weather analysis toolkit has been provided," added Levermore. "This will provide data that can be manipulated in design discussions with the client."

"We have also provided more information to deal with semi-extreme periods, and data which acknowledges that buildings are being designed with a greater degree of thermal inertia."

Weather and solar data

Cold and warm weather data for eight UK locations is now included, namely Belfast, Birmingham, Cardiff, Edinburgh, Glasgow, London, Manchester and Plymouth. A new test reference year has also been included, as have heating and cooling design temperatures for a range of European locations.

The existing weather data for 1957-1966 has been replaced by data for the period 1976-1995, which is not only more relevant but more detailed. This updated data also reflects the inexorable increase in mean annual temperature. Note that both 1976 and 1995 had exceptionally hot summers, so the weather data for those particular years will be very useful for testing the limits of a natural ventilation design.

Much of the information for the eight geographical locations is in tabular form, with binned frequencies of 24 h and 48 h average temperatures. It is interesting to note that while the existing Guide A gives a 1% exceedence for 30°C (September 1976), the new data gives a 2·5% exceedence – a significant percentage increase.

We should be pushing for a level of airtightness...and making it clear to the client and the architect that our design is based on that’

Chris Twinn

Whereas the existing Section A2 has tables of dry and wet bulb temperatures, the updated edition will enable designers to plot the percentage frequencies of combinations of hourly dry and wet bulb temperatures on a psychometric chart so that specific enthalpy exceeds can be identified.

A new part in A2 deals with climate change. The four scenarios given by the UK Climate Impacts programme are included, plus the predicted temperature changes for south-east England and central Scotland for the 2020s, 2050s and 2080s, indexed against known data for the period 1961-90.

Section A2 will include real weather data in sequences of one, three, five and seven day periods. This data can either be stitched into a test reference year, or be used "as is".

Very extensive wind data has been included for selected locations in Europe and the UK (also to be provided on CD-ROM), although this is meant only for general guidance as the data may only be appropriate to the area local to the meteorological station.

Providing good hourly solar data was a problem for the Section A2 authors as the Meteorological Office measures little in the way of solar radiation – important for determining summer conditions. Although the Meteorological Office has provided some data, it is likely to impose a levy for its use. The CIBSE is currently in discussion over a suitable sum.

Section A3: Thermal response and plant sizing

Section A3 of the CIBSE Guide is a straight revision with updates and improvements as deemed necessary by changes in technology and tighter International, European and British standards. Hence the sections covering thermally bridged constructions, solid and suspended ground floors, basements and windows have been substantially rewritten.

Comprehensive tables on the thermal conductivity, absorptivity, density and specific heat properties of a wide range of building materials have also been added to Section A3 of the draft Guide. Assembled by the University of Strathclyde, these tables will help designers calculate the thermal admittance of various forms of construction with much greater accuracy.

Heat loss calculations for opaque structures have been improved from those in the 1986 version, with the two methods – proportional area and combined – rationalised and backed up by examples.

Although the proportional area method is now enshrined in the Building Regulations, section author Peter Owens said that, in extreme cases, it can produce U-values up to 5% lower than they should be. "The combined method is much more accurate," he said, "and the CIBSE Guide gives examples for the designer including cavity block, cavity timber stud and tile hung walls."

Mistakes within the U-value calculations for solid ground floor construction have also been removed. "The existing Section A3 assumes external temperatures more suited to walls and roofs," said Owens, "when in reality the temperature swings are more likely to be seasonal than diurnal." The section on windows incorporates a new calculation procedure. This covers multiple panes, low emissivity glass and gas-filled glazing. The section also includes data on blinds and curtains.

Section A5: Thermal properties of building structures

Compared to the other sections in draft Guide A, Section A5 – the calculation of heating and cooling loads and heating plant capacity – is not very different from the 1986 revision. However, it now incorporates Sections A8 and A9 to provide more integrated guidance and eliminate inconsistencies – "of which there were a great many," said principal author, Mike Holmes.

Junior members of the profession might be tempted to treat the relative humidity figures as rigid limits and thereby install humidification unnecessarily

Brian Mark

A major objective of the revision was to introduce computer-based methods of determining thermal response without losing the virtues of the manual thermal admittance procedures. Designers will be free to use professional judgement as to which method is appropriate for the design application.

Guidance is given on the calculation of steady-state heat losses, the determination of summertime temperatures for naturally and mechanically ventilated buildings, and the calculation of peak space cooling loads.

The various methods cover both steady-state and dynamic calculation techniques, with steady-state calculations relying on:

The full model is designed to represent radiant flows between surfaces in an arbitrary-shaped space, combined with general convention heat transfer and one-dimensional heat flow through walls. This is a manual model, as a complete interpretation of the dynamic response of a building is not "currently practicable".

The basic model is a computer method which uses radiosity and longwave radiative exchange equations for a six-sided enclosure. The simple model is a manual method of calculating the heat transfer of surfaces, based on average and typical properties. The reference model is a computer method based upon the fundamentals of longwave radiant exchange between surfaces.

Inevitably there will be differences in the answers from the reference, basic and simple models, but their importance all depends on the context.

The CIBSE's approach to non-steady state models has been fraught with debate2, not least because there are many commercial dynamic computer programs already on the market. To avoid arguments, Section A5 will merely list the essential attributes of a dynamic simulation program, and leave it up to the individual designer to decide which simulation tool is most appropriate.

It is essential that the heat flows in and out of a structure are correctly represented, hence a simulation program should include representations of:

"However, the presence of these features in a dynamic thermal model does not mean that the resulting model will be valid," said Mike Holmes. "For example, the different moisture absorption characteristics of building materials can significantly affect system loads."

Essentially, the Guide stipulates that a basic (dynamic) model should:

Holmes stressed that designers will need to test computerised simulation models against these criteria. For example, a simple test could be conducted comparing a building with heavyweight thermal mass against one without thermal mass. The lightweight building should then follow the sine wave of the weather data.

Solar gain factors

One of the problems with assessing cooling loads is calculating the solar gain factor through glazing at different times of the day. Section A5 will include two methods of doing this: the mean solar gain factor and the alternating solar gain factor.

It is vital that air cannot circulate in cavities around insulation

Max Fordham

Whereas calculating mean solar gains is relatively easy, calculating alternating solar gains is more difficult. One reason is it usually lags the solar intensity by between zero and two hours, the duration of the lag depending on surface factors and the amount of solar shading. High surface factors give rise to delays of about 60 minutes, while low surface factors give delays of about two hours.

The cooling load tables in Section A5 are based on hourly calculations, the effect of the angle on incidence onto glazing, and solar data based on EC recognised algorithms.

Workshop discussion

Inevitably for a publication which has spent the last 12 years in revision, certain parts of it were considered out of step with contemporary thinking, issues of fabric airtightness and building structure being cases in point. "We can calculate the fabric losses to the nth degree," said Gibb's David Wood, "but we don't seem to be able to go much further on unplanned infiltration. Ventilation through cracks and between components is not dealt with anywhere in the Guide."

The BSRIA's Gay Lawrence-Race strongly agreed. "A lot of information in the Guide leads us to believe that accurate calculations will improve the product, but there are lots of other variables. Real U-values are dependent on how things are installed on site.

"If we can't be accurate on site, perhaps what we need to do is match system performance closer to how buildings really perform," she concluded.

Chris Twinn responded, saying that designers must be aware of the differences between site and design but should not assume the lowest common denominator. "We should be pushing for a level of airtightness and making it clear to the client and the architect that our design is based on that," he said.

The CIBSE's policy on environmental criteria was generally regarded as reasonable, although delegates questioned the 40% and 70% relative humidity levels quoted in Section A1. Section author Paul Appleby admitted that the wording needed "a bit more thought" to prevent building designers misunderstanding the context.

The section on ventilation for smoking environments also alarmed many delegates, mostly because of ASHRAE's problems with Standard 62.1989, but section author Paul Appleby argued that the CIBSE differs from the ASHRAE in its responsibilities for setting guidance for healthy and comfortable indoor environments.

Delegates questioned the basis 16 litres/s/person for "average" levels of smoking, (average being defined by Paul Appleby as 20% of people in the room smoking at any one time). "I would suggest an amendment saying that 'this is a suggested value and does not necessarily indicate comfortable and healthy conditions'," ventured Chris Twinn.

The Guide's presumption that clients would become more involved in design iterations worried the building physicists at the workshop, Whitby & Bird's Duncan Price calling it "wishful thinking" that better weather data will incline more clients to sit down and discuss risk analysis with the consultant. Chris Twinn wryly suggested a full page advert in the Financial Times to alert developers that they will be "grilled" on the new weather data.

Delegates were very interested in thermal bridging, particularly for curtain wall construction where problems with bridging are legion. This can now be calculated more accurately, although designers will be expected to obtain manufacturers technical specifications.

One delegate complained that vented cavities in curtain walling are a "confidence trick" by curtain walling manufacturers to circumvent the Building Regulations. "Calculating the thermal performance of that construction can be very risky," he said.

Phil Jones from the University of Wales argued that better advice was required on the location of insulation in ventilated cavities. "If you put insulation on the outside of the ventilated cavity, you get the same U-values as if you put it on the inside," he said, "but actually the thermal performance is different because the heat short circuits into the ventilated cavity. Architects do this a lot," added Jones.

Max Fordham argued that the next edition of the Guide should include guidance on achieving the insulation values that are calculated. "It is vital that air cannot circulate in cavities around insulation," said Fordham. "Few construction people realise that, and few of us tell them to button the insulation up. The Guide should give that advice."

Delegates accepted the introduction of noise criteria, but Mike Silcock was concerned at the NR 35 requirement for offices. "NR 35 seems a very onerous requirement," he said. "Perhaps we should add a rider that states how you measure it and the circumstances under which you conduct the test."