Use of The Metalcasting Metrics Data Spreadsheet

The Data Factors for Metrics Estimating page details tonnage by alloy family; by molding process.  Tonnage is categorized by tons shipped, tons produced, and tons melted.  The three subcategories begin with tons shipped, and the additional two are derived from typical internal + external[1] scrap rates and mold yield (See footnote 2.)  Melt loss is in addition to “Tons Melted” (See footnote 2) because it can be analyzed separately, where applicable.  Also on this page is a further breakdown of the metalcasting tons produced by alloy family; by molding process.  The units of this breakdown are “Million BTU’s per Ton Produced,” and subcategories of these units are Melting and Other Processing.  “Melting” correlates to 1) Furnace melt energy saved per pound of metal in castings poured; 2) Casting melt energy saved per pound of metal in castings shipped; in the definitions on page one.  “Other Processing” correlates to 3) Non-melting energy savings, page two.  Finally, there are conversion factors for BTU’s to Dollars for five common energy sources.

The extension of “Melting” and “Other Processing” BTU’s per ton produced by “2003 Est Tons Produced” results in energy consumption by alloy family/process for the metalcasting industry.  The total of 329.9 Million BTU’s is similar to the U.S. Department of Energy’s Metal Casting Industry of the Future (IOF) estimate for total metalcasting energy consumption of captive and non-captive foundries.  Looking at the breakdown among “Melting” and the elements of  “Other Processing,” the 56.3%/43.7% split is close to the Metal Casting IOF’s published energy consumption split of 55% Melting and 45% Other Processing.⁶

ESTIMATING TECHNIQUES

A common problem in developing metrics for R&D projects is lack of sufficient quantitative data.  Each of the three examples below suffers from this problem.  Example three has more extensive difficulty from lack of data.  When data is thin, the following techniques are helpful in making reasonable estimates:

• Make a diligent effort to discover the quantitative data that does exist.

• Further perspective on thin data can be obtained quickly via web searches.  Sometimes websites provide new data or helpful qualitative information.

• Educated guesses of experts or people directly involved in the process(es) which the R&D is addressing.

• When making your own “not-so-educated” guesses, break the guesses down into the smallest elements practical, based on the available data.

• The thought process required in breaking down thin data into smaller elements often provides insights that helps improve the guess.  For example,

  • Mentally segment the industry to which the R&D applies as finely as possible:

  • Are there segments of the industry that would receive the technology differently?

  • Are there segments of the industry to which the technology would not apply?

• Estimating the smaller elements helps identify opposing factors or factors that move in the same direction, but at different levels that would be missed in a simpler “macro” guess.

• Use a baseline for estimating differences between an existing technology and a new technology

  • It’s tempting to make estimates of the difference between an existing technology and a new technology by claiming “percent reductions.”  Those are quick and relatively easy to do when data is thin.

  • Instead, estimate the baseline metric(s) of the existing technology, using the above “breakdown” techniques.”

  • Then estimate the same metric(s) with the impact of the proposed new technology

  • Subtract the new technology metric(s) from the existing technology metric(s) to determine the estimated savings.

  • This approach requires more disciplined thought than the “estimated difference” approach, and it facilitates more of the above “breakdown techniques.”

• Do a sensitivity analysis on the estimate

  • Put the estimate in the context of any available overall industry or similar technology data

  • Does it seem reasonable?

  • If possible, make a new estimate from a different perspective or a different set of assumptions.  Does the new estimate offer any insights, any affirmation, any doubts?

  • Cross reference reasonableness against other benchmarks

• Avoid estimating changes in industry structure as a means of increasing a metric estimate.  For example, in the metalcasting industry there have been major shifts from ferrous metals with high energy content in melting to light non-ferrous metals with much lower energy content in melting.  Molding process improvements that further reduce the energy content in the light metal castings has driven some of these shifts.  These shifts have been the result of R&D, much of which has been funded by the Department of Energy.
  
  Therefore, it is tempting to forecast further or similar shifts in the industry segments.  Those shifts can have dramatic increasing effects on metrics.
  
  However, those shifts are driven by factors beyond the scope of a proposed R&D project.  They depend mostly on the reaction of markets that the metalcasting industry serves.  These shifts go beyond “adoption rate” estimates for a new technology; they are driven by the demand from the metalcasting industry’s customer base.  If metrics are based in whole or in part on these downstream, demand-pull assumptions, the basis of comparison by the funding organization can become muddled.  It is recommended that metrics estimates be made in the context of a “snapshot” of the present industry segment distribution.