Holdfast Training Services Trial Blog

Pre-fabrication and off-site manufacture seem to be very popular in the UK construction industry at the moment and I have lucky to have been the lead on the largest pre-fab install on my project – a 132m long, 4m wide services culvert linking the two buildings currently being constructed.

The install consists of 21, 6m long pre-fabricated modules with 12 modules installed on side of a central vertical shaft and 9 on the other.  The modules carry chilled water, high-grade hot water, low-grade hot water, mains cold water, boosted cold water, sprinkler supply, HV, LV, ELV, Fire Alarms and Fibre-Optic Data.  Water is transported through a series of cast pipes varying in size from 1″ to 16″ and the pipes are connected using Teekay Couples with a fixed point in the centre and expansion bellows either end of the tunnel to allowing for thermal expansion.  All pipes are lagged and seated on slip rings.  The cabling is fixed to a variety of containment (ladder racking and trunking) as shown in the pictures.

Background

During the procurement process, a study was conducted to consider the use of prefabricated modules in the tunnel rather than traditional on-site MEP install using Unistrut bracketry.  This study estimated the cost to traditionally fit out the tunnel to be £745,757.  The study concluded that there could be a saving by using pre-fabricated modules manufactured off-site.  Since my project uses BIM Level 2, the required module was subsequently modelled in 3D and inserted into the consolidated building model.  In my opinion, BIM significantly reduces the design risk when designing complex modules or bespoke structures and for us the ability to model each module in 3D and then move it through a virtual 3D tunnel using the as -built information from a laser scan was invaluable and actually led to the height of the modules being reduced as it was clashing with a low point in the tunnel.

Once the design had been agreed and approved by the Principle Designers (BDP), it was sent to Skanska Fabrications for a mock-up to be constructed.

As the install lead I visited the mock-up a few times and was able to provide feedback that led to design changes that would make the install easier (one such improvement was to set back the low level Unistrut that forms the walkway to give better access to the pipe connections and Teekay couples – you can see the issue in the above photo).

Once the design had been amended and agreed upon, the package to manufacture and deliver the 21 modules was put out to tender and awarded to Balfour Beatty Engineering Services (BBES) for a sum of £460,805.  Although this looks like a significant saving on the £745,757 estimated cost of traditional install, other costs need to be factored in before a true comparison can be made.  In addition to the manufacture fees, the installation costs (which I tendered under my own package) totalled £86,325, the cost of realigning and connecting the pipe work was agreed at a cost of £55,493 and connecting the electrical containment has been quoted for at a cost of £11,500.  This brings the total cost of manufacture and installation to £614,123 which is about 18% less than the estimation of the traditional install and has been hailed as a significant saving to Client.  This cost does not take into account the additional cost of designing a more complex system and comparing the costs to an estimation is unreliable as the costs could have been over-estimated.  Therefore I believe the actual saving, if indeed they was any, is actually quite small especially since a traditional install would have procured through a 2-stage competitive tendering which would likely result in a contract sum that is less than the estimation.

One of the greatest advantages of pre-fabrication is that you can remove it from the critical path in the programme and reduce the time required for on site install.  By manufacturing off site you can manufacture early and store the products rather than having to wait for an area to become free or for interfacing/preparatory work to be completed.  However what I have found in reality is that the time to design a pre-fabricated solution is much greater (especially if modelled in 3D using BIM) and so although you may plan to construct early, manufacturing can be delayed by the design and drawing production which is what has happened on my site.  In my case, the manufacture was delayed due to a lack of drawings (this is general issue on my project!) and this in turn meant that the tunnel was left empty and ready for the modules while the modules were still be constructed meaning that there was no real time saving. Infact, if we had gone down the route of traditional install then we would have fitted out the tunnel earlier than we did.

A disadvantage of pre-fabrication is that the design needs to be agreed and fixed earlier in the programme which then creates a greater risk from design changes later on.

Installation

The tunnel modules were installed as follows:

First of all a rail was installed by my sub-contractor using a jig-system:

Then the modules were dropped down the central shaft using a crane and dropped onto the rails:

They were then dragged into position using a winch:

Unfortunately the installation did not go as smoothly as the above description and I had to make changes both to the method and the state in which the modules were delivered.  Despite this, we installed the first module on 24 Nov 16 and the last module on 12 Jan 17.  Now we have to finally align and then bolt together all of the modules before we can realign and connect the pipework and containment.  Then we can also take a site measure and construct the ‘make-up section that sits in the centre of the tunnel and will fix the pipework to the tunnel wall (required for expansion).

Snagging

Inspection of the installed modules has highlighted several issues.  The first is that most of the pipework is out of alignment and so will need moving in order to be connected.  The mechanical sub-contractor is arguing that the amount moving and aligning required is beyond the scope of their contract and so are requesting an instruction and extra cash.  It is my belief that the pipes have moved during transportation due to cyclic loads.  The pipes are on slip guides and the slip guides are fixed to the modules using Unistrut channel nuts (zebedees) which rely on friction to provide a fixing.  Therefore in theory all of the pipes could move if forces are applied to it outside of the design conditions.  A lot of the containment is also out of alignment.

Another issue is that some of the modules do not fit together as some of the walkways and kick plates stick out too far.  This makes me question how the modules were jigged at the factory – a concern I have raised with BBES.  However the greatest issue with the install is that some modules are sitting noticeably higher than the others.  I am still investigating the reason for this but believe it can be only one of two possibilities:

1. The rail is high in places which could be due to a poor install or imperfections in the tunnel itself -several were picked up on the laser scan!
2. Some modules were incorrectly constructed or mis-jigged.

Either way I need to come up with a solution and so I am looking at ways to reduce the overall module height.

Given the variety of snags and issues, I have assembled a multi-trade team using 4 sub-contractors that will go through each module and adjust every pipe, rack and section of trunking to allow all the modules to be moved as close together as possible which will minimise the gap between the pipework (maximum allowed gap is 8mm).  This team will start next week which gives me a few days to work out how to lower some of the modules!

Conclusion and Recommendations

It is my experience that the conventional understanding that using pre-fabrication saves time and reduces programme risk whilst increasing costs is incorrect since on my site we have seen the exact opposite.  So far on Project Laureate, using pre-fab has led to on-site delays (with knock on effects on other packages) and has cost less (18% less) than a traditional install. However, it is worth noting that this could be unique to this site due to the scale and complexity of this install or the selection of materials.  Also the cost of the traditional install is reliant on an estimation which may not be accurate.

Prior to conducting this task it was my understanding and the understanding of my colleagues that the modules would arrive on site and line up perfectly, reducing the amount of time required for the on-site fit-out.  Again this has not proved to be the case as none of the modules fit together without remedial works and extra time is now required for realignment.  I would urge anyone running a similar task in the future to assume that any bespoke modular system being delivered to site will need a certain amount of realignment and adjustment on-site.  At the very least this requirement should be priced as a provisional sum during tendering and should be included in the programme.  This is particularly important for large and complex structures or systems with several variables (e.g imperfections in the final structure or multiple connections).

During the design of a module or pre-fab structure, assume that the structure it will be installed in is riddled with imperfections and plan against this to reduce the risk of delays and additional costs during the install.  This could be done by adding a means of adjustment in each axis.  Although this will incur additional costs, in the long it will likely save money.

When considering loads on the structure or system, also consider the temporary state such as transportation to reduce the risk of structures deforming or equipment shifting.  This could lead to the requirement for bespoke bracing or supports that be will used during delivery to site.

Regularly visit and ask your sub-contractors to visit the factory during manufacture to provide feedback and to facilitate design changes that will make the install easier.

Although this package has proven to be an engineering challenge and is my greatest headache at the moment, would I use off-site manufacture again? Yes I would as I still think if done right it can offer either cost savings or time savings.  However, I would implement the points above and would most likely tender it out as a design and build package to reduce the risk of delays due to an overstretched Project CAD Team. If using a Project CAD Team, I would recommend only considering the use of pre-fab for complex installs if BIM is the primary method of design.

Mark, can I count this as TMR 4?!