At the Fall BIMForum/BuildCon conference in Dallas two presentations about actual prefabrication experience - for data centres and for large automotive plants - both mentioned shortage of skilled labour as an important motivation for turning to prefabrication for construction projects. In addition to reducing onsite labour requirements, Improved quality, greater safety, faster project delivery, and in some cases reduced costs were mentioned as the key benefits of prefab.
A compelling presentation on the market drivers for and benefits of prefabrication for data centre infrastructure was given by Matthew Englert, Vice President at Rosendin Electric and President, Modular Power Solutions and Chris Crosby, CEO of Compass Datacenters. Data centers, which used to be defined in terms of servers and racks of servers, are now defined by containers of servers. Timing is of the essence for building data centres. Compass advertises "We’ll deliver your dedicated data center in six months from breaking ground on a pad ready site or we’ll pay you $100,000."
The electric power requirements of data centres are reckoned in the hundred of kilowatts and even megawatts. A significant portion of the cost of a data centre is the electric power rooms to convert and manage the power for the servers and other equipment in data centres.
Both Chris and Matthew said that an important motivation for turning to prefabrication for the power rooms is the decline in the number and quality of the skilled tradesmen required to build power rooms. Older and more highly skilled trades people are retiring and they are not being replaced by younger people. This is partly a North American problem because there are limited apprenticeship programs. Perhaps because in the past skilled trades people were simply imported from Europe.
There are other advantages to prefabrication. It is easier to attract skilled workers to work in a plant under a roof with the amenities of factory environment as opposed to working outside often in inclement weather. Weather is always a risk on a construction site. Another advantage of a factory environment is safety. Statistics show that injuries are less in a prefab environment compared to a construction site. Quality is another factor that is easier to control in a prefab plant.
For data centre power rooms the prefab "skids" built by Modular Power Systems are huge weighing 60,000 to 80,000 pounds and requiring large flatbed trucks to deliver from the prefab plant to the construction site. You can watch the construction of a couple of data centres at the Compass site and the power rooms are prefabbed offsite and delivered on flatbed trucks. They are offloaded onto concrete pads that have been prepared in advance. The process is very efficient and reduces the congestion of trades people at the construction site.
Chris was blunt in his assessment of the appropriateness of existing BIM software for designing data centres. His perspective is that BIM software is no better than traditional CAD. The problem is that the level of detail (LOD) currently supported by BIM software is simply not detailed enough. He argues that if you are going to use 3D design software then mechanical design tools such as Solidworks are better than BIM tools. He believes that BIM software is ripe for disruption. Designing buildings and other infrastructure in the future is going to require new design tools and that is going to require significant investment in new technology. This is a very similar sentiment to Greg Luth's perspective on the advantages of high definition building information modeling (HD BIM) for structural engineering of large manufacturing plants such as the Tesla Gigafactory in Nevada.
Based on his experience Chris believes that prefab always costs more the first time. Repetition is required to bring the costs down to where prefab is cheaper than stick built. Therefore Chris and Matthew are convinced that the next step is to move beyond prefab to manufacturing.
General Motors came to similar conclusions based on their experience in using Epsilon Industries prefabbed utility system upgrades for four GM automotive plants. As presented by Jeffrey Johnson, Manager, Process Energy Initiatives, Bob Stevenson, Senior Vice President, Ghafari Associates, Chris Wiederick, President, Epsilon Industries, and Christopher Hofe, Project Director, Barton Malow Company, the problems that encouraged the GM team to prefab most of the utility components were a shortage of skilled trades and wanting to avoid congestion resulting from many trades tripping over each other at construction sites. They estimated that prefabbing reduced their costs by about 20% and enabled them to deliver completed utility plant for the automotive factories in eight months instead of eighteen. For example, a complete chilled water system including a safety system and weighing 85 000 lbs was manufactured in Kingston, Ontario, trucked down and and was up and running in 2.5 weeks. Also since quality assurance for the chiller was completed in the factory in Kingston, the handover was virtually instantaneous compared to the prolonged period required for a stick built chiller.
For GM shortage of labour is a major problem and was a major motivator of the decision to use prefab components from Epsilon Industries. The GM team repeatedly emphasized that they didn't know how they could have staffed stick build projects at these four sites. Onsite congestion at a construction site is also a major problem that is relieved by prefabbing components. Work conditions on a construction site are often poor because of bad weather. Weather is always a risk that can hold up a stick built project. They also reported that supervising in a factory environment is much easier compared to a construction site. GM found it easier to recruit labour for a factory environment and they was able to retain better quality labour, people who wanted to settle down, as opposed to the high proportion of transients typical of construction projects.