3.7 ENGINEERING COST MODELS
3.7.3 How to estimate costs using engineering methods
3.7.3.1 Estimation of capital expenses
According to Ruhmer (1991:138), the intended use will determine the amount of money spent on any estimate of capital costs. This will determine the method of estimation and its probable accuracy. Various procedures are available for preparing estimates, ranging from rapid and low-cost ones to more
87
formal and tedious detailed design ones. The total capital cost, CTC, of a project consists of the fixed capital cost, CFC, plus the working capital costs, CWC, plus the cost of land and other nondepreciable costs, CL:
CTC = CFC + CWC + CL
The accuracy of the estimate depends on the degree of project definition and the fact that project definition is generally vague at the start, improving as the project study and design progress.
Ruhmer (1991:138) classifies estimates in four classes in ascending order of accuracy. The class I estimate establishes the “order of magnitude” only. A class II is known as a “preliminary estimate” and has a tolerance of -15% to +25%. A class III estimate is termed a “definite estimate” with a tolerance of -10% to +10%. A class IV estimate is known as a “detailed or revised estimate”
with a tolerance of -5% to +5%. A class IV estimate cannot be made until 40 to 60% of the actual construction work on the project has been completed. The four classes of estimate and the information required to compile them are illustrated in tables 3.6 and 3.7 below.
88 Table 3.6 Capital estimate classifications
Estimate definition
Estimate description
Target accuracy Required information
Class I Order of magnitude
Dependent on available data;
range to be indicated on estimate
General site conditions Plant and infrastructure layout
Process flow figures Timing of project for escalation
Class II Preliminary -15% to +25% As for class I plus:
Major equipment Specification
Preliminary piping and installation diagrams based on research and
development guidelines Actual plant and
infrastructure location and layouts
Class III Definitive -10% to +10% Complete process design Engineering design and layout complete
Detailed design 20 to 40%
complete
Project construction schedule
Construction contract Class IV Detailed or
revised
-5% to +5% As for class III with detailed engineering design
essentially complete Source:
Adapted from Ruhmer (1991:138)
89
Table 3.7 Information required to compile estimates Estimation
definition
Basis for estimate Major
equipment
Other materials Labour
Class I By comparison with similar work done before, with adjustment for capacity, escalation and site conditions Class II Budget
prices/and or recent purchase costs, including freight, adjusted to current rates
By ratio to major equipment costs
Labour material ratios for similar work, adjusted for site
conditions and using current rates
Class III Firm quotations with critical items committed
Firm unit-cost quotations (or current billing costs) based on specified
quantity take-off
Estimated man- hour units (including productivity assessment) using expected labour rate for each job classification Class IV As for class III
with most items committed.
As for class II with material on approximately firm basis
As for class III – some actual field labour
productivity may be available Source:
Adapted from Ruhmer (1991:139)
Other authors have used different labels and different boundaries for the various types of estimates. Ahuja and Walsh (1983:45-46) refer to planning estimates, preliminary engineering estimates, detailed engineering estimates and construction phase estimates. According to Humphreys and Wellman
90
(1996:7), AACE International (formerly the American Association of Cost Engineers) has proposed three classifications of cost estimates. The major types of estimates proposed by AACE, in increasing order of accuracy, are provided in table 3.8:
Table 3.8 AACE capital estimate classifications
Type Accuracy Order of magnitude
Budget Definitive
-30% to +50%
-15% to +30%
-5% to +15%
Source:
Adapted from Humphreys and Wellman (1996:7)
For the purposes of this study, the estimate classifications used by Ruhmer (1991:139) will be used as guidelines, because this is a South African publication widely used in the industry - for example, Mintek, the developer of the publication, as well as Kumba Resources.
The design, and hence the cost of the plant, depend on the skill and points of view of the design engineers for the project. To cover up errors of judgment, the overdesign of control or safety factors may be used. Many design engineers are reluctant to practise intelligent risk taking.
Factors that tend to increase the capital cost of a plant, include the following:
overprovision for safety
overprovision for standby equipment
unnecessarily robust supporting structures
building enclosures
inclusion of nonstandard-sized equipment
expensive construction materials
Design and process economics are interrelated, and without adequate cost estimates, a satisfactory design is not possible. To avoid uneconomic
91
commitments, management should be supplied with adequate cost estimates so that they can make the necessary decisions. The main steps in preparing a capital cost estimate are as follows (Ruhmer 1991:140-141):
The initial idea of the process and the specifications of the size and type of operation are formulated.
Physical and chemical data are collected from the literature or from laboratory experiments.
A preliminary flow sheet, incorporating the required unit operations, and showing the main items of equipment, is prepared.
Heat and mass balances are prepared.
Specifications are drawn up of temperatures and pressures at various points on the equipment flow sheet.
Design calculations are made to select and size the main items of equipment such as pumps, heat exchangers, reactors, columns and tanks.
A list of the major process equipment is compiled, including details such as size, capacity, materials of construction pressures and temperatures.
The cost data are collected from vendors.
The total delivered cost of all the items of major process equipment is estimated. The purchase cost, quoted by the supplier, may be multiplied by a factor of 1,03 to provide the approximate delivered cost.
For order of magnitude and study estimates, the simplest method of estimating the fixed capital cost of a plant, based on design information, is the Lang factor method. The fixed cost is given by
∑
CEQL FC f C
where CFC is the fixed cost; fL is the Lang factor, with a value of 3,10 for solids processing, 3,63 for mixed solids-fluids processing and 4,74 for fluid processing; and CEQ is the cost of a major item of process equipment. The original Lang factors were developed in 1948 and were based on a limited range of statistical information. The method is most effective in companies that
92
use the Lang factors based on their own company data. Attempts have been made to improve upon this basic approach but with little success. Although many of these attempts are logical in their approach, it is not possible to ascertain the absolute value of each subfactor for South African conditions because most of the published data are for the USA (Kharbanda & Stallworthy 1988:23; Ruhmer 1991:141).
Another component of capital expenses is working capital which calculated from the following items (Ruhmer, 1991:142):
inventory of raw materials and supplies - one month’s supply (consumption at purchased value)
inventory of products and in-process material - one month’s production valued at manufacturing costs)
accounts receivable (or extended credit) - one month’s production at sales value
available cash – one month’s manufacturing expenses