Confusion with Woods cost estimation methodology

[sasa000]

Using Woods 'Rules of Thumb in Engineering Practice' (2007) Appendix D cost estimation approach (easily available to download online as pdf - won't link as probably not kosher).

Section D.2 / Page 377 suggests there are two ways to estimate the Bare Module (BM) cost correcting for L+M*, however I seem to be calculating wildly different (L+M* - FOB) costs for each (ie. the BM installation cost not including actual equipment)

These are the two approaches, ignoring for now other factors like pressure correction factors:

i) Reduce the reference L+M* factor (typically at CS) by the value/plot suggested after correcting for Material of Construction => BM = FOBcs x Alloy Factor x L+M*ref x (L+M* correction factor)

ii) Calculate the BM cost as CS (FOBcs x L+M*ref), then add on the differential BOM cost for the MOC eg for stainless steel. BM = (FOBcs x L+M*ref) +(FOBss - FOBcs)

If we're just looking at the L+M part of the costs, these two approaches give wildly different answers:

Taking for example a radial blade centrifugal fan as an example (Appendix D Section 2.1), say at the reference 10Nm3/s condition (and ignoring CEPCI correction for now):

FOBref = $15,850
L+M*ref = 1.4-1.7 (assume 1.4)
Alloy Factor (CS) = 1.00
SS = 2.5 -> gives correction factor of around 0.65 per Figure D.1

Method 1: BM cost = $15,850 x 2.5 x (1.4 x 0.65) = $36,058
The L+M* part of said cost therefore = BM cost - FOB cost = $36,058 - 15,850 x 2.5 = -$3,567
So therefore the L+M* Material correction actually makes the BM cost cheaper vs reference case (L+M -1).

Method 2: BM cost = $15,850 x 1.4 + $15,850 x (2.5 - 1) = $22190 + $23775 = $45,965
The L+M* cost is therefore just the CS L+M* cost (1.4 - 1) x $15,850 = $6,340

Woods seems like a fairly respected methodology, but the two methods suggested give wildly different answers! As far as I can tell I'm using the correct approach (though the text could do with a few more developed examples). I wonder if method 1 is supposed to be written to state that the L+M* correction should only be applied to the incremental material corrected FOB cost, but the text is not written to define it that way and that still wouldn't solve method 2 giving a negative correction (ie. therefore stating that a higher material grade installation cost would be cheaper than CS).

 

[1503-44]

No problem with posting links. Proprietary copies are a different thing.

Most of these thumb rule costing methods are for larger aggregate plant cost estimates, not for any one specific equipment item. They are notoriously not suitable to use straight out of the box and normally require serious fine tuning. Especially in today's economic climate.



--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[shvet]

Pricing is a challangeing thing. Many real projects I have been involved in had equipment where pricing had been broken and let's say SS was cheaper than CS.
Your problem is you suppose that (a) equipment/unit price is derived from technical parameters and (b) cost approximations are developed based on a wide range of industry applications. No and no respectively.

In your particular case cost equation has been developed based on a couple of independent estimations or real references which became not relevant to a real pricing.

Appropriate pricing/costing is extremely expensive and provides huge benefits to a user over its rivals/competitors. Do you believe such info would be readily available in a published free download source? Or it is a sort of 'second-hand' no one wanted to pay for and therefore got to a published source?

 

[pierreick]

Hi,
Consider this resource to support your work.
Don't ask more from an estimate.
Pierre

 

[sasa000]

Thanks everyone, I appreciate that cost estimates should be treated with caution. My question was more around why the same source appears to have two methods that are so significantly inconsistent. These costs wouldn't be used in anything more stringent than Class 5 (or Class 4 at a pinch if appropriate quote can't be found in time)

I *think* the correction factor given should be based on ignoring the FOB cost, so in fact should be applied to (L+Mref - 1) x (L+M* correction factor), certainly this gives a more consistent resultant (eg. SS selection factor shouldn't give a lower L+M* than CS now, it should be similar)