Correction Factors in Cost Estimation for Bidding

Software Cost Estimation and Sizing Methods - RAND Corporation

Or, just use a spread sheet.

A spreadsheet should be good enough for this, but there are subtleties:

V-groove welds will cost about proportional to the square of material thickness; J-grooves will be less, say t^1.5?
Cost will vary stepwise with length, not perfectly linear, but with a jump for each course of plate.
Somewhere in your scheme, put multiplier(s) for spot-market material pricing, labor rate changes, etc.

Just some quick ideas. I've done this before, but for a more limited range of sizes than you are contemplating. And don't forget specialties like impact testing, radiography, heat treating, etc.

Many thanks. I agree with you. What I am proposing to do is to identify, through analysis, a representative or equivalent vessel - with a representative Diameter, Length, thickness, material. We would start with a base estimate of labour hours, for this equivalent job. (Material cost, labour cost can be corrected as applicable at the time of actual estimate).

I would then like to involve our team (to avail of organizational memory and knowledge) to agree, through vigorous discussion, on multiplying factors to indicate impact of following additional parameters.

Number of Nozzles

Quantity and type of externals and internals

(As you suggested), Number of more-than-normal inspection, testing or client witness points.

Volume to be produced of same design.

Where produced - inside the shop or in the open yard. If latter, we may have to add corrections to productivity and quality impacted by external climate - especially in a place like Saudi Arabia.

Finally, risks. Risk to schedule compliances, cost budget compliance, and, hence to profits.

After applying these factors we should be able to arrive at an estimate of labour hours for the new job. And, using actual past data, we can validate and fine-tune these factors.

You were mentioning that you have done this kind of an exercise before. Can you provide more details

My old Excel sheet for this is on an off-line laptop, so I would have to do a lot of fiddling to get it to you. I doubt that it would help you much, because I specialized it to 250 psi and for a small range of ammonia refrigeration vessels.

I entered the diameter and length, and let the spreadsheet calculate the thickness and number of plate pieces (or preformed heads), compute the weld material used and time, etc. Then I had a prompting list of standard nozzles and simply entered the diameter and number of each (inlet, outlet, instrument column connections, etc.), as applicable.

In my scheme, I priced each vessel separately, rather than a generic type/size with corrections.

You should base your spread sheet on two statistical analysis:

comparison between post-calculations and

estimations for bids.

According to the production you mention a sufficient data base should be available. You can even consider in your estimations the price evolution and labor cost trends which can be quite good estimated for reasonable delays between bid and production dates.

I was years ago confronted with a similar problem for high pressure rams and the post calculations available allowed me to develop a simple correlation between some design parameters and final cost of manufactured product. It was an exponential equation and the uncertainty was quite low. It was possible to compensate it by a safety coefficient.

As an example you could consider dimensions and mass as basic parameters since dimensions and weldings are correlated in a rough manner. Complexity can be a 3rd parameter. A multi variable correlation could lead to a simple way for a rough estimation and comparisons with past calculations could deliver the uncertainty and define a compensation safety coefficient.