Thursday, 20 September 2012

System Building turned on its head!


It’s no secret that one of my favourite television programmes is “Grand Designs”. My favourite episode is “Ben Law’s house in the woods”. If you are not in the UK, then this may not make that much sense to you.

It’s also no secret that, as an engineer, I like system built construction, but I don’t particularly like system built houses. So far, the ones on “Grand Designs” which have impressed me are: “The Huf House” (http://www.huf-haus.com/en/home.html) and “The ADApta-house” (http://www.adaptahaus.co.uk/).  After last night’s new episode (19th September 2012 on UK Channel 4) I need to add another one. I’m afraid I don’t know the name of the system yet, but I’m going to find out. I also don’t usually enthuse about television programmes, but this one is worth watching, even if you don’t particularly like Grand Designs and Kevin McCloud (http://www.channel4.com/programmes/grand-designs/4od).

Description

If you haven’t seen it yet, here’s my summary and thoughts:

The clients are a pair of lady university professors in a civil partnership. They’re obviously approaching retirement. What is admirable is that they decided to act as patrons to a small industrial design firm with a big idea. I don’t especially like the house, but I do like the method of construction. The house is very “Modernist” and the method of construction suits that kind of design. The method reminds me, very much, of techniques suggested for constructing things on the Moon or Mars!

The designers had a “big” idea, but had never built a house before. This was going to be their prototype. The deal that the clients and the designer had was that the clients paid for the construction, but the “thinking” and “error-fixing” was done for free by the design team. The design team did the construction themselves. The site was awash with degrees of various kinds, and lots of enthusiasm.

Most system-building relies on using standard modules which are constructed off-site. Often this means “the bigger the better”. That way most of the work is done in the controlled environment of the factory and all that happens on site is the final assembly of the modules. In many cases the modules are fully fitted out before they leave the factory. Great on site, but it does mean that the client has to make all the decisions before construction starts and that may be months in advance.

This new technique stands all this on its head (or “inside out”, or whatever metaphor you like). Everything is made on site!

 Let me describe the process for this particular house.

The foundations were screw-piles, topped by an engineered wood framework (timber “I” beams) and a plywood base floor. So far, so ordinary.

The “factory” arrived at site inside a steel container with a dropdown side. All it needed to get going was some concrete slabs to stand on and an electricity supply. Even that could have been a mobile generator. 

The raw material was sheets of 8x4 construction plywood. The house was built of semi-standard modules. The factory used a computer controlled milling cutter to cut a sheet into the components for an enclosed box. The box was assembled by the workers as the components were finished. It typically took 4 minutes (give or take) to assemble a box and fasten it together with a nail gun. Each box was (usually) constructed from a single sheet of ply and was therefore of such as size and weight as to be easily handled by 2 people.

Once a box was completed it was carried to the house and fastened into place. Each box is numbered by the mill (number routed into the surface) and manufactured as it is required. The assembly drawings tell the workers where it has to go. The boxes were fastened together using wooden staples made by the mill, from the same plywood sheet as the box components.

This was a two storey building, and once the shell was finished, it was roofed, rendered and clad fairly conventionally.  Plumbing and electrical first fix was done conventionally but made very convenient by having the routing for the services built into the boxes. Windows and doors are subcontracted and fitted into openings built into the shell. Once the building is watertight and secure the boxes are filled with blown recycled newspaper insulation (so that’s a foot of insulation on all walls, and the floor and (I think) the roof).

The show followed the standard Grand Designs format and identified problems on the way through, but actually, especially considering it was such an experimental construction method I didn’t see that many.

Problems and Vulnerabilities:

Like all building using precision, factory built, components, you are very dependent on the quality of the foundations and initial setting out. It has to be square and level because there is no opportunity to adjust things later. This wasn’t a problem on this occasion because whoever did the groundwork knew what they were doing (probably not the designers of the house).

Before the first floor was up there was some heavy rain and the designers were concerned that the dampness might mess up the dimensions of the boxes. It doesn’t seem to have been a problem in practice.

On the ground floor, the modules turned out to be “a bit of a tight fit” in places. This was because the designer had not allowed for a proper fitting clearance. It wasn’t too severe a problem, and was rectified for the first floor by an adjustment to the cutting design. Probably achieved by a single parameter being adjusted, or alternatively, by re-dimensioning the first floor boxes (if done that way, it probably turned into an all-nighter, which we didn’t see).

The windows were a perfect fit in their allotted places, but the client didn’t like the proposed fastening method. Actually, neither did I; screwing through the frames even when the screw heads are covered with caps, seems a bit crude for such a high spec building. The alternative (much better) involved hidden brackets, and required some manual chiselling to make rebates for the brackets. The designer said he would adopt the new attachment method for future buildings and the rebates would be routed into the components by the factory.

The flat roof didn’t have quite the correct fall to drain properly. The additional drain had to be modified to satisfy the building inspector that it had sufficient capacity. Once again, this was a problem with experience and detailing, and the solution has now been accepted into the design parameters used by the designer.

One of the clients fitted the blanking plates for the holes used to install the insulation herself. She said they were too tight a fit. Again, a detailing and dimensioning problem with can be solved for future constructions.

The clients felt that at times the designer took too much control of the design. They said that they felt “they should have been more assertive” at times. They also said that they felt that part of the deal was to give over some control to the designer, and that they were entirely happy with the result. This seems to have been a good client/patron, maker relationship.

The project ran slightly over budget (15% ?), but don’t they all?    

Strengths

  • What I like about this method is the way that everything is made on site. There is considerable flexibility.
  • It seems to mix good bits from traditional building: first fix and fitting out on site, inside a weather-tight shell.
  • The client is not committed to anything until the actual cutting starts. I wouldn’t recommend making changes at the last minute, but they would be possible.
  • The client is not committed to decorative finishes until the shell is up and they can walk around it. This solves the problem of people visualising spaces. It also creates the possibility of adjusting the finish to suit the budget.  
  • Stock and material control is easier. There is one main material: sheets of 8x4 ply. Building component boxes (which do not have to be all the same) are made as they are required. The bill of materials from the design program (don’t know what it was) handles the quantities required. Made components are numbered by the factory to match the construction drawings, so identification should be straight forward (no looking for “right handed hoople-grommet”) and components are constructed just before they are needed for construction, so there is no stock-yard to search through.
  • The “factory” is just the contents of the shipping container, plus a covered assembly area on site. No need to pay for the costs of a large factory area when it isn’t working.

Limitations

  • The method lends itself to boxy “modernist” construction. All right if you like that sort of thing.
  • I suspect that the 8x4 sheet imposes some limitations, like the maximum height of ceilings (think about it). Even this can probably be got round, with imagination. Similarly, total number of storeys may be limited. None of this is at all serious. I would regard it mainly as “characteristics” of the method and the material.
  • It is probably easiest if the factory is building parts for one building at a time. That may represent a limitation on “utilisation efficiency”, but frankly I don’t see it as too serious. If you have two or more buildings on the same site (they don’t have to be identical, just make them in sequence, and maybe move the factory to each one as you build it).

Overall

 I like it! I may not like the product, but the clients did, and that is what matters. I can see this carving a niche for itself. Similarly, I think the idea of moving the precision factory onto the site, making custom components just-in-time and then assembling them as required is very interesting. The added-value chain might be interesting too (sheet of ply, component box, house shell, finished house etc). 

No comments: