(English) Onsite First Time Assembly Works: Debunking the Most Common Myths
I have to admit, that the first time I gave much thought to Onsite First Time Assembly (OFTA) for new machines was following a dinner conversation with a very experienced engineer from one of the large Italian underground contractors. The engineer suggested that it was his opinion that on large diameter machines, perhaps above 8 m or so, that much time and money could be saved by an OFTA program, which would allow the machine to be initially assembled on location. He said he did not believe the huge amount of labor expended to completely assemble a TBM in the shop, which then had to be repeated in the field, could be justified. He argued that labor was being duplicated and the result was a longer delivery period that was reasonably required. This conversation, taking place in the mid-nineties, is what first triggered my thinking on the subject. Well, that conversation and couple of glasses of Brunello di Montalcino! My engineering friend raised a very valid point and I’ve given the subject much thought since, and discussed OFTA with many people in the industry. It appears to me that, even though the method has been successfully used on many of our projects worldwide, there are a few “myths” that have gained traction regarding the risks of OFTA. I’ll address two of these OFTA myths.
Myth No. 1 – It will be a disaster if the parts don’t fit together on the job site! That is why a full factory assembly is required.
This myth doesn’t take into account three factors:
1. Partial assembly: Major sub-assemblies are pre-assembled in the factory. The main bearing and seal assembly, for example, is fully factory assembled. It is only the major parts that may not be pre-assembled. However, in many cases, we pre-fit the pieces; forward shield to outer telescopic shield, outer to inner telescopic, inner telescopic to gripper shield, and so on for a double shield machine. So the part fit-up is checked but the entire machine is not put together in the factory.
2. Modern measurement devices: In those cases where it is not possible to pre-fit two pieces in the factory, we can use modern Coordinate Measuring Machines (CMMs) or “Laser Trackers” to take precise measurements of both pieces to insure fit up when they meet in the field.
3. In-field repairs: When an offshore oil rig has a component failure do they disassemble it and take it to the mainland for repairs? When repairs are required on hydroelectric turbines and generators, do they always take the big parts to a machine shop to repair? No, they do not. Many repairs of large scale equipment are made in situ, wherever the plant is located. If it is discovered at the job site that a component has been mis-machined / manufactured it is most likely a minor error (remember the CMM measurements and component fit-ups in factory partial assembly) and repair can be effected on the piece in place at the job site. In Robbins experience, this has always been the case and such repairs have been carried out when discovered on site and without impact to final startup schedule.
Myth No. 2 – The labor cost on an underground job site is far higher than labor cost in a factory. It will cost far more money and time to assemble the machine on site for the first time.
OFTA Myth No. 2 leads to a wrong conclusion regarding cost, and fails to examine fully the potential benefits.
1. Cost: This above statement is generally true; however, it does not necessarily follow from this statement that money will be saved by having a factory assembly plus a job site assembly. Robbins history with in-factory and OFTA assembly reveals the following:
Traditional Factory Assembly
- Full factory assembly hours: X hours
- On-site assembly of a fully factory assembled machine: 0.5X hours
- Partial factory assembly for OFTA delivery: 0.5X hours
- OFTA site assembly of a partial factory assembled machine: 0.7X hours
Where the value of X is dependent upon the size and type of TBM as well as the complexity of the backup system.
If we assume that the cost of labor on the job site is twice the cost in the factory, then we can use $100/ hour for the job site and $50 / hour for the factory. The total cost for the two methods is then:
- Traditional method cost = X hours ($50/hour) + 0.5X hours ($100/hour) = $100X
- OFTA method cost = 0.5X hours ($50/hour) + 0.7X hours ($100/hour) = $95X
In short, the total costs are nearly the same, or perhaps with some savings in favor of the OFTA method.
2. Schedule: Another flaw with Myth No. 2 is that it does not consider a large potential benefit: the savings in schedule possible with the OFTA method. By not completely assembling the TBM in the factory, it is possible to deliver a working TBM at the job site one to two and a half months earlier than with a traditional full-factory assembly and delivery. This is a significant savings for most projects, when the site assembly can be done at this early stage.
3. Training of site personnel: Another potential benefit that Myth No. 2 does not address is that of training. The contractor’s site personnel who are involved with OFTA assembly and testing get far more training due to the additional hours spent on the assembly and testing, and the larger Robbins crew present to assist and advise. Robbins’ experience indicates that contractors who opt for OFTA deliveries are frequently capable of taking over the full operation and maintenance of their new TBM much quicker than contractors who opt for a traditional delivery. This is due to the much deeper knowledge the contractor’s personnel gain during the OFTA assembly and testing.
In summary, the question to be asked when making the decision to use OFTA or go for a traditional factory assembly is: What are the potential risks and what are the potential rewards? One must examine all potentials of the OFTA scenario, pros and cons, to come to the correct solution. I’m not claiming that an OFTA scenario is correct for every project. For example, if the start of boring of the tunnel is not on the critical path for the project and the TBM is a smaller diameter unit, it might be advisable to allow a full factory assembly. That way, when the job site is finally ready for the machine, it can be assembled a bit quicker on site. Again, if the start of boring is not on the critical path and the labor cost difference between job site and factory is larger (e.g., job site is in Finland, TBM factory is in China), then it could be that a full factory assembly can be justified on a cost only basis. However, in nearly every case when the start of boring the tunnel is on the critical path, then the faster delivery possible by OFTA is clearly favorable.
The biggest impediment to more widespread use of OFTA is limited thinking: looking at the potential risks without looking at the potential rewards. Tunneling itself is fraught with risk, yet contractors take on jobs everyday due to the potential rewards. OFTA is deserving of a similar analysis.
About the Author
Joe Roby (B.S., Mechanical Engineering, University of Washington) has worked in the tunneling industry for more than 20 years. He started at The Robbins Company as a stress analyst specializing in finite element analysis of complex structures. Subsequently he was a member of the 19-inch cutter development team. For five years he was managing director of Robbins refurbished and leased TBM division. He has authored many technical papers for conferences and industry publications on subjects ranging from cutters to TBM assembly and rebuilding practices. Today he serves as Robbins’ Vice President – Production & Logistics.