Lower waste, lower labour, lower costs…are the theoretical benefits of prefabrication realistic? Alison Luke reports from an on-site trial.
ff-site prefabrication has been trumpeted as the way forward for construction for several years now. The benefits and savings to be made by using this form of installation are said to be substantial, but it has been difficult to prove the claims. This is set to change.

Crown House Engineering (CHE), one of the biggest promoters and appliers of the technique, aims to find out the facts by doing a full-scale comparative trial on a project in London's Savoy Place.

The company has been applying off-site prefabrication since the mid-1990s. It is also heavily involved in research being carried out in partnership with Nottingham and Loughborough Universities. But now it is taking the research to a practical level.

The former Shell-Mex House on the north bank of the River Thames is the site of this experiment. A complete refurbishment of floors 5-9 within the nine-storey building was to be carried out prior to occupation by the new leaser, international media group Pearson. And timescales were tight.

The first fix m&e contract was won by Crown House on a negotiated fixed price tender. The contract involved the installation of ductwork, electrical containment trays and the heating and chilled water pipework. The client specified the services and timescales required. "How the system was put in," explained CHE project manager Brian Skinner, "was down to us. And the job leant itself to prefabrication."

As well as the fast programme, each floor will have a similar purpose, with pc workstations dotted throughout. The floors all measure around 3500 m2, are largely open plan, with only supporting columns breaking the space, and the servicing was to be split into four similar zones on each – a client choice for control reasons. This mirroring of the floors meant that much of the installation was the same throughout, making prefabrication ideal. But Crown House saw an extra opportunity for the project.

"Heavy production demands" on CHE's Wolverhampton manufacturing centre at the time the building's modules were to be built offered a chance to do something different. "The company needed to start benchmarking," stated Skinner. And with the similarities between the floors, detailed and realistic comparisons of traditional and offsite prefabrication methods were possible.

Drawings began in late August 2000 and work on site at the start of October, with completion set as 18 December. At the design stage floors 5-7 were chosen for prefabrication, while installation on floor 8 was to be by traditional methods. "To stay within time," explained Skinner, "we couldn't afford to do more than one floor that way." The installation included 25 km of pipework, 8 km of traywork and 4 km of ductwork.

Differences between the two methods were obvious on the floors during installation. Cutting and bending machines sat in one area of floor 8, with various green waste bins in another. Inbetween, piles of pipework, bends and joints sat neatly but bulkily. So much material was required on site that to increase space on floor 8, some had to be stored on one of the other floors.

Aside from this materials storage, floors 5-7 appeared clean and empty. Coordination of module delivery was carefully done to ensure on-site storage was minimised and schedules kept. A logistics engineer was employed to tie in the deliveries from Wolverhampton with transfer onto the floors and installation.

Even if the project hadn't needed this level of coordination, the site itself made it necessary. One 2 m-wide alley provided the only access point, causing delivery and size issues, and the only storage space was on the floors. The modules were further limited by the size of the mechanical hoist used to transfer them to the floors – the final size was 1.5 x 0.5 m.

There are around 100 modules on each floor. Although similar, each had its own drawing showing its exact requirements. Slight changes had to be made on site to some of the modules: where they met the existing service risers and joins were too complicated to detail on paper.

Each module was supplied with pipework in place and brackets for the ductwork and the electrical containment. The ductwork arrived separately and was simply slotted into place. "It did not suit to include the ductwork in the modules," explained Skinner, "it was quicker for it to be constructed elsewhere at the same time as the modules were being made in Wolverhampton," showing how a mix of the methods may prove the best option.

The inclusion of the cable trays leaves an easy job for the installation of the electric cabling on the prefabricated floors; the electricians will simply have to lay cabling on the supplied system. On the traditional floor however, much more work is required to complete the same task: brackets and supports must be put in individually, requiring more labour to maintain the same timescales, and producing more on-site waste.

To ensure like-for-like comparisons the labour force was split into two teams, one completing two modular floors and the second doing a floor using each method. "The general consensus among the workers on site was that the modular system is a lot easier," reported Skinner. It also required less labour.

On the prefabricated floors around six staff were needed, but on the traditional floor around 15 were needed to install the same services. Overall the onsite workforce was around 40-45; over 100 workers would have been needed to do all the work the traditional way.

This reduction in numbers brings health and safety as well as financial benefits. "Doing the job in a controlled factory environment is a lot safer," stresses Skinner. Work can be completed quicker and at safe and convenient angles and distances, with awkward, perhaps overhead, work eliminated for all but the final fixings on-site. In theory, with less staff, the chances of accidents will lower.

Also reduced was the number and type of facilities such as toilets, drying rooms and canteens needed. With less staff to provide for, fewer facilities are required to meet the regulations.

Waste was found to be an important factor. The costs of the services was less using off-site manufacturing, due mainly to the controlled environment in which the modules are produced and subsequent lack of waste. There is also the issue of removing waste from site: "There's a lot of materials...you can't just throw in a skip," reminds Skinner, "A lot of aftercare and paperwork is involved."

The job has been monitored throughout and the results will be used for marketing and tendering purposes. "We need hard facts that say to clients, 'this is what a modular system does' and 'this is what a traditional system does'," says Skinner. The Pearson House project seems a good first step to finding those facts.

Related files/tables