About / Publications / Publication

Office Form, Energy and Land Use


By Dr. Dean Hawkes, Richard MacCormac, Dr. Francis Duffy, Paul Townsend & John Herbert

This essay first appeared in RIBA Journal June 1978

In this joint article, Dr Dean Hawkes, of the Martin Centre at Cambridge, examines the energy implications of the office
form which started as a speculation in the firm of MacCormac and Jamieson, and the urban and future design potential of the idea are described by, Richard MacCormac, who, with David Prichard, has developed some working studies. Dr Francis Duffy has acted as office layout consultant. Paul Townsend, of the Martin Centre provided costings and John Herbert of Land Use Consultants provided
landscape information.

This a hypothesis about the design of office buildings which started from an enthusiasm for the quality of rooflit courts, ambiguous inside/outside spaces, such as that in Wright’s Larkin building,or in Aalto’s Rautatalo building, in which the convivial activities of the city, shops and cafes are introduced into the sheltered environment of the surrounding offices. Such buildings propose a kind of urban space as apt for northern climates as the piazza is for the south.

This qualitative appreciation of courts is supported by quantitative attributes of density and environmental behaviour. The studies outlined here are a further investigation of the primary theme of architectural research at Cambridge – the comparison of alternative built forms. They specifically address what might be called the ‘classical’ comparisons between court and pavilion forms of building. Earlier studies (1) have demonstrated the advantages of the court in terms of the efficient use of land and these have had effect in the field of housing layout (2) but whereas for housing the court yard array, poses problems at the inter sections, for offices the intersections offer a location with vertical circulation and services. As with housing, the concept offers the potential of dual land use, sustaining the plot ratio of office accommodation while offering the courts for public use.

Here, the potential of the court as an efficient user of energy is the main subject of examination and forms a critique of the conventional wisdom that the most efficient built form, thermally, is that which is ‘more cube-like’, continuing earlier work at Cambridge (3,4)

The present work opens with a speculation. The question is can we achieve a balance between the variables of the built form which would offer both the inherent economies of a daylit and naturally ventilated building and the low external surface area of a deep-plan arrangement? The study begins with an investigation of an array of glazed courts, surrounded by a band of daylit space, and concludes with two worked examples, one rural and one urban.

Investigation of a Generic Form

Initial studies concentrated on a single court and the space associated with it. First, the daylighting question was examined. Studies using scale models showed that adequate daylight for most office activities would be achieved in most of the building during the summer months. These considerations determined the basic dimensions of the court. Conversion of measured daylight
factors into absolute illuminance indicated that levels exceeding 100 lux would be achieved 5m from the windows on the lowest floors at the beginning and end of the working day in the months of April and September.

Next, the summertime thermal conditions were studied by estimating mathematically, by the BRE Admittance Method (5) the temperatures which were likely to be reached in both the court and office space on a hot sunny day. The results showed that the temperature in both spaces would be acceptable provided that most direct solar gain could be effectively excluded from the
court. This could be achieved by a variety of means and detailed analysis showed that a ‘reversed north-light’ configuration would offer an acceptable compromise between solar control and the admission of daylight. With this, the peak temperature in the court would be of the order of 25°C and in the top floor offices less than 24°C.

At this stage, the form could be seen to offer a satisfactory visual and thermal environment in the summer months without the use of substantial energy consuming systems. Having established this the next step was to calculate its winter energy consumption. These calculations were based on a ‘standard’ court and its associated office space. It was assumed that the office roof was well
insulated (U = 0.5W/m²°C) and that the glazed court roof was double glazed (U = 3.0W/ m²°C) and that the walls between the office space and the court are single glazed (U = 5.6W/ m²°C) The temperature difference between indoors and outdoors was taken as 20°C and a ventilation rate of 5 air changes per hour was assumed. The analysis showed that the temperature in the court resulting from heat loss to it from the offices would be of the order of 13°C and that the heating requirement, under steady state conditions and allowing for occupancy gains would be 35.5W/m².These calculations did not make any allowance for the effects of heat gains from
lighting or machinery, nor did they allow for the benefits of passive solar gain in the winter months.

The next step was to make an estimate of annual energy consumption using the ‘degree-day’ method (6). This showed that, for locations in the London area, with the building occupied eight hours a day for five days a week, the heat requirement would be a little over 230MJ/m² per annum. After making an allowance for a small amount of energy consumed by mechanical ven tilation the total would be of the order of 250MJ/m² per annum.

To put this figure in perspective, atypical figure for a ‘conventional’ air-conditioned office building would be 1 300MJ/m² per annum and, for recent deep-plan offices of the Integrated Environmental Design type typical energy consumption is around 800MJ/m² per annum (8).

Towards application

Most buildings will, of course, be restricted, by site constraints and the brief, to a small number of courts. In these cases, the edge condition will  inevitably increase the energy consumption of the building as a whole.

The worst case must be that of the single court surrounded by a further 7m wideband of office space. The behaviour of this outer band is taken to be like that of a notional double banked corridor of offices in a conventional slab block with 50% glazing in the external walls, double glazed (U = 3.0 W/m²°C) and well insulated wall construction (U = 0.5W/ m²°C) which calculations show as being
likely to have an annual energy consumption of about 625MJ/m² per annum. The effect of an outer band, with these characteristics, is to raise the annual energy consumption of the single court to around 500MJ/m²per annum. All other cases would be superior to this and analysis for the two and four court forms showed that they would consume approximately 450MJ/m² per annum and 400MJ/m²
per annum respectively.

These results show the economies of scale which one would expect from those forms which maximise court-side office space relative to perimeter space. It is possible, however, to attend to the problem of the building for a small site by the use of a simple tactical device. If the generic court were to be completely enclosed by a windowless wall with good insulation (U = 0.5 W/ m²°C) – a ‘teacosy’ – it would offer annual consumption of approximately 300MJ/m² per annum.

The result endows the whole system with enormous potential by allowing the designer to choose appropriate combinations of generic courts with or without perimeter space to suit any particular set of circumstances whilst, at the same time, preserving part of the energy conserving properties of the generic court. A small number of the possibilities based.on four courts are illustrated. For a
glimpse of the immense combinatorial possibilities beyond this, the reader should see the work on The Animals of Architecture by March and Matela (9).

In essence, the environment in these buildings would be characterised by its variability, within limits, and by the ability of the occupants to exercise a degree of control over their own local conditions. A fluorescent lighting installation would give general background lighting after dark and each workplace would have its own individually controlled task lighting. The narrow cross section makes it unnecessary to use complex air-handling systems for heating, with their associated service zones, a simple hot water heating system with radiators or convector units would be adequate and would, again, allow a degree of local control. Mechanical ventilation would be required to provide permanent extraction from the service cores at the junctions between courts and this would be used to provide the primary wintertime ventilation. Air would be drawn from outside, through the court, into the office spaces and out through
the cores. This arrangement would mean that incoming air would be pre warmed by its passage into and through the court before entering the office spaces. In the summer months, the mechanical ventilation of the cores would continue, but the office space and courts would rely more on natural ventilation through opening lights and controllable openings at the edge of the court rooflights.

Rural study

The lack of external constraint in this hypothetical open site allows this example to be generalised and its main characteristics are shared by the second example on an urban site. It might represent the headquarters of a large private or public organisation accommodating up to 5000 personnel in a total area of 70 000m².

Implicit in the scheme is an attitude to the recent evolution of office building design and in particular to the two acute problems facing office planners in the aftermath of burolandschaft – firstly to retain the freedom of the open plan and yet create identifiable places within it and, secondly, to invent building forms which have the capacity to be either highly cellular or, to a high degree, ope.
Hertzberger showed the way to solving the first problem in Centraal Beheer. The concept illustrated here reflects the same concerns as Hertzberger but has the potential to solve the second problem. Office space of relatively narrow depth which is punctuated by courts and held together by a grid of circulation can offer cells, small areas and deeper spaces five or six workplaces deep, provided it is possible to shift the bands of circulation from one side of the office space to the other, so the scheme can offer advantages of planning
flexibility quite apart from any environmental considerations.

At ground level, the courts are connected by a mall which offers the primary circulation route through the building. Given the population, this would generate some of the characteristics of an urban street relating to shared amenities such as restaurants and shops or creches. A principle of this scheme is that no office space is more than one court from the edge and that accommodation
looks out as well as in. In energy terms, it lies between the theoretical cases illustrated in 7b + c, that is about 425MJ/m² per annum.

Urban study

This study takes as its context a half hectare redevelopment site in Covent Garden. The crucial test here is that the concept offers a commercial plot ratio. A scheme consisting of four courts with only single banking of accommodation around the outside offers a plot ratio of about 4:1 in five storeys and about 5:1 in six storey construction. The generic form has been distorted to suit these conditions while maintaining its essential advantages.

The single banking of accommodation around the outside allows the scheme to be heavily ‘wrapped’ in insulation on its road elevations, with limited glazing for views, while admitting daylight and ventilation to the working areas from the internal courts. Consequently, the building gains the maximum benefit from its court arrangement and its energy consumption would be superior to the 300MJIm² per annum of the case illustrated by 7d.Assuming a building with a total floor area of 18000m the annual energy costs can be estimated as, being about £30000 which might represent a saving at current costs of £60 000 per annum over a totally air conditioned building consuming 1000MJ/m² per annum.

Apart from the performance of its office environment, the particular interest of this example lies in the contribution which it makes to the repertoire of urban space and form. While maintaining its high plot ratio, it offers to the community the same kind of sheltered space as the traditional glazed markets of Covent Garden. Assuming basement car parking, it opens up a proportion of the site to the
public for use as gar dens or as access to ground level restaurants and shops. As an urban form, it preserves traditional urban scale and proposes a planning convention that is the reverse of that which has recently dominated and destroyed the structure of our towns. For instead of isolating buildings from one another as pavilions surrounded by draughty channels of open space, it suggests an opposite stereotype of building against a notional party wall or site boundary of given height to create useful sheltered space within the site and
continuity of urban form without.

Future research and development

The purpose of these initial studies has been to test a hypothesis. The results show that a built form based upon a relationship between a glazed court and adjacent daylit, naturally ventilated space offers the prospect of creating a very acceptable office environment at an annual energy consumption considerably lower than that achieved in much recent construction. The worked examples demonstrate the way in which the theoretical studies may be used as a spring board for realistic designs for particular circumstances.

In the future, the potential of these studies may be extended both theoretically and in application. Further theoretical analysis would place precise limits upon the dimensions of the generic form within which its inherent performance characteristics are maintained. Combinatorial studies would spell out explicitly the relationships between energy consumption and an extensive set of forms based upon the generic case. These would also provide the basis for detailed investigation of the relationship between the built form and
capital, costs and costs in use.

The application of these studies is not restricted to questions of office building design alone. The underlying principles and performance characteristics would hold when applied to any other building type whose dimensions and functional organisation related to those of the generic form. Educational buildings, libraries, shops, some industrial uses and some health buildings come to mind as likely subjects.


1 Martin, L. and March, L. Land use and built forms, in Cambridge Research, April 1966.

2 MacCormac, R. Housing form and land use: new research, RIBA Journal, November 1973.

3 March. L. Elementary models of built forms, in Martin, L. and March, L. (Eds.) ‘Urban space and structures’, Cambridge University Press: Cambridge, 1972.

4 Hawkes, D. Modelling the environmental performance of built forms, in March, L. (Ed.) ‘The architecture of form’, Cambridge University Press: Cambridge, 1976.

5 This method is outlined in the ‘IHVE Guide1970, Book A’.

6 These calculations were based on the data presented in the ‘IHVE Guide 1970, Book B’.

7 Milbank, N.O. Energy Consumption and Cost in two large air-conditioned buildings, ‘IHVE/BRS symposium: thermal environment in modern buildings’. Building Research Station Current Paper 40/68.

8 This figure is derived from data giving the measured annual energy consumption of two IED buildings.

9 March, L. and Matela, R. The animals of architecture: some census results on N-omino populations for N-6, 7, 8. Environment
and Planning B
, 1974, Vol. 1, p. 193-216.