The cost of accuracy

In general, improving performance will come at the cost of accuracy. The benefit of using a given "trick" to reduce computational time lies thus in the eye of the beholder—and the system of interest. 

Designing an experimental procedure involves selecting an appropriate method of analysis based on established criteria, such as accuracy, precision, sensitivity, and detection limit the urgency with which results are needed the cost of a single analysis the number of samples to be analyzed and the amount of sample available for... 

W. R. Park investigated the cost and accuracy of estimates for a projec t with a total cost of 1 million as shown in Fig. 9-41 Co.st Engi-... 

The Dickson N method [R. A. Dickson, Chem. Eng., 54, 121-123 (November 1947)] is a variation of the detailed price takeoff. Various circiiits for each type of pipe are completely priced for a base size. Another chart gives an N faclor for all other pipe sizes. Multiplying the cost of the circuit for the base size by the appropriate N factor yields the estimated cost of the new circuit of the desired pipe size. The method depends for its accuracy on periodic repricing of the base-size circiiits in order to keep the base charts up to date. 

Availability of large digital computers has made possible rigorous solutions of equilibrium-stage models for multicomponent, multistage distillation-type columns to an exactness limited only by the accuracy of the phase equilibrium and enthalpy data utilized. Time and cost requirements for obtaining such solutions are very low compared with the cost of manual solutions. Methods are available that can accurately solve almost any type of distillation-type problem quickly and efficiently. The material presented here covers, in some... 

The accuracy of a CFD program is greatly influenced by the number of cells in the grid. The larger the number of cells, the more accurate the solution. Additionally, the cost of computer hardware and calculation time combined with the accuracy of a solution depends on the preciseness of the grid. 

A key aspect in most cost estimation studies is the evalutUion of the cost of individual pieces of equipment. In addition to quotations from equipment manufacturers, there are several methods for developing cost estimates. For a typical conceptual design, it is normally acceptable to develop an order of magnitude or preliminary estimate whose accuracy is within -h/— 25 to 50%. The following methods may be employed to develop order of magnitude estimates ... 

The cost of preparing an estimate increases from about 0.1 per cent of the total project cost for 30 per cent accuracy, to about 2 per cent for a detailed estimate with an accuracy of 5 per cent. 

Miller claims that if the basic equipment estimate has an accuracy of 10%, the most likely plant estimate should have an accuracy of 14%. This is much better than the ratio or Lang estimates, and considerably more accurate than Nichols said was possible with this type of data. 12 He claimed that there is a direct correlation between the cost of an estimate and its probable accuracy. Ever since he stated this in 1951, cost engineers have been trying to prove him wrong. 

The cost of a unit is then obtained from Figure 9-4. To obtain the cost of the plant this figure needs to be multiplied by the number of units, a factor (1.33) to account for utilities and general facilities, and the current or projected value of the Engineering News Record Chemical Cost Index. The accuracy of this method is unknown but should be better than the Lang factor and not as good as Miller s. 

These estimates are for a national distribution and will increase depending upon the number of subpopulation groups for which frequency distribution data is desired. Determining the sample size, therefore, requires decisions regarding the percentile of interest and level of accuracy needed, and balancing these against the costs of the required sample. 

Similarly, improvement in the accuracy of the nuclear dynamics would be fruitful. While in this review we have shown that, in the absence of any approximations beyond the use of a finite basis set, the multiple spawning treatment of the nuclear dynamics can border on numerically exact for model systems with up to 24 degrees of freedom, we certainly do not claim this for the ab initio applications presented here. In principle, we can carry out sequences of calculations with larger and larger nuclear basis sets in order to demonstrate that experimentally observable quantities have converged. In the context of AIMS, the cost of the electronic structure calculations precludes systematic studies of this convergence behavior for molecules with more than a few atoms. A similar situation obtains in time-independent quantum chemistry—the only reliable way to determine the accuracy of a particular calculation is to perform a sequence of... 

Air quality measuring systems that detect for example the leading substance C02 (refer to chapter 53.3.3) do not require the high accuracy of expensive measuring systems. Therefore, not much effort needs to be put in the development and construction of C02 detectors. In general the accuracy of such systems amounts to 10%, which is achievable for the cost of an electrochemical cell. Fast measurements aren t needed but nevertheless averaging to increase accuracy is possible and recommended. 

---