Are you visiting from outside your region? Visit your regional site for more relevant services and pricing.

Comparing British office workers journeys-to-workSeptember 2008

Kate Fewson1 and Adrain Leaman2 examine eight British office buildings to see how they compare in terms of CO2 contributions made by journeys-to-work.

So-called green buildings may or may not achieve their energy targets (most don't), and may or may not be liked by their occupants (although many are), but what about the carbon dioxide emissions resulting from journeys made to and from those buildings?

The authors have carried out a preliminary study of existing data in order to establish some yardsticks and to get some idea of the orders of difference.

Table 1: the eight buildings that formed the study in carbon dioxide emissions from journeys to work.

The study buildings are a subset from Building Use Studies' (BUS) UK database3. These have had building-user surveys carried out on them which include a journey-to-work section. The journey-to-work analysis was originally introduced when it became clear that many buildings were increasingly dependent on recruiting staff to service them, many of whom travel longer distances to work because they could not afford to live close by.

The approach was originally developed by Kate Fewson on three study buildings4, then extended by a further five buildings for this article. Altogether, there are data for 21 offices, 11 higher education buildings and 11 schools, but here we are only looking at eight of the offices.

Table 2: the results summary for the eight study buildings.

The offices were selected to cover a mixture of types and locations, ranging from large city-centre offices to smaller buildings in semi-rural and urban fringe/ring-road settings in England and Wales. Some of them have green building credentials, and some are occupied by organisations with an avowedly green agenda.

Note that the sample is not big enough to allow robust comparisons between these types and locations. At this stage we need to understand orders of difference and magnitude.

The BUS survey asks people how they get to the study building, whether by their own car, a shared car, a bus, train, or bicycle and so on. People know quite accurately their best and normal journey times, and especially recall the worst cases.

Table 3: average normal journey to work times for the study building

As the BUS questionnaire asks for respondents' home postcodes, and the postcode of the study building is known, Google maps can be used to obtain approximate travel distances to and from the study buildings. The task is laborious, but less so than other methods, including that of asking the respondents themselves, who tend to be very poor at estimating distances.

Knowing the journey mode and the distance travelled enables the use of published carbon dioxide coefficients to estimate emissions5. For example, for cars (as make, model, age and fuel type is not part of the BUS survey) a published (2005) ratio6 of 82:18 petrol:diesel is used for the consumption of cars in the UK, along with 0.348 kg carbon dioxide per mile for petrol and 0.295 kg CO2 per mile for diesel. This gives an estimate of 0.338 kg CO2 per mile, with assumed average consumptions of 30 and 40 miles per gallon respectively.

Obviously, it would be better to use exact figures derived from detailed observations and monitoring of journeys, congestion, fuel use and other variables, so that the estimations could be improved, but we do not have that information. In its absence, we have to be as clear as possible about the assumptions, so that the reader can judge our estimates for themselves. To this end the summary of our assumptions are available from the Usable Buildings Trust website.

The most striking differences include a six-fold difference between the lowest CO2 contribution (Building E, with over 80 percent walking or cycling) and the highest (Buildings A and H, both with over 85 percent car-based commuting).

Building D travel to work plots by journey mode and distance

City-centre buildings (B and G, with higher proportions of train and bus travel, 52 percent and 63 percent respectively) have about half the carbon dioxide contribution of car-dependent buildings.
Buildings occupied by organisations with an explicit environmental agenda (C and D) do better despite locational disadvantages. Staff in Building C make more of an effort. Where the organisation has introduced green travel plans (as in building D), car use has been cut down from a possible 90 percent to an actual 50 percent, of which 17 percent are car shares.

Figures 1 and 2 are journey-to-work plots for buildings E and D by journey mode and distance. Building E is the best performing in this study, with an unusually high proportion of people walking and cycling. Building D has introduced green travel plans, with enforced car sharing and buses to help cut down car-borne emissions. In spite of this, one employee commuted weekly by air.

Building E travel to work plots by journey mode and distance

The lowest average journeys to work (20 minutes) is again building D. As might be expected, Building G, the London case, has the highest average, 55 minutes: this equates to nearly two hours commuting on a normal day, taking into account journeys both ways. Congestion and delays can make this much worse.

The average worse case journey time for building G is 93 minutes (not shown in table 3). If this is divided into the average best case time (47 minutes), this gives a congestion index for the building: a score of 1.97. If all the journey times were the same every day, the index would be 1, the best possible case. The congestion index for building D with its low contribution is 29/19, a score of 1.53.

What it all means

So what does this tell us? It shows that the study buildings with over 80 percent of people commuting to work by car are up to six times worse for commuting-related carbon dioxide emissions than those where 80 percent walk or cycle. It also tells us that buildings with occupiers with a committed green agenda are likely to get commuting-related emissions down by a half, even if they have locational problems, such as being located on an out-of-town business park with relatively poor public transport where the temptation to drive is overwhelming.

However, the study is only based on eight cases, so it important that these results should be seen only as a first stab at understanding orders of magnitude and difference. The next step is to compare commuting emissions alongside building emissions, in order to look at their relative contributions. There is little point in being painstaking about reducing building emissions through better design and management if these are cancelled out or swamped by emissions from travel and the downsides of congestion.

If long commuting times mean that it becomes even harder to recruit staff to run buildings efficiently, especially in city centres, then this is a further nail in the carbon coffin.


1 Kate Fewson is with Design Group 3 Architects.
2 Adrian Leaman is with Building Use Studies and The Usable Buildings Trust.
3 Further details about the Building Use Studies survey method may be accessed on the Usable Buildings Trust website.
4 MSc Architecture, Advanced Environmental and Energy Studies, Centre for Alternative Technology, Machynlleth
5 Hampton, D. (2006) The Edge Pledge: Carbon Audit Excel Spreadsheet. Accessed 31 May 2007 and 27 April 2008
Hampton, D. (2007) MSN Carbon Emissions Calculator: Assumptions and calculations. Accessed 11 June 2007 (Thanks to Dave Hampton.)
6 A further useful source on carbon calculators is Botterill C, Internet-based tools for behaviour change, Paper presented at sEuropean Council for Energy Efficient Economies. Summer Study 2007 Dynamics of Consumption Session 9.