Purpose of, and Approach to, Approximate Estimating
The primary function of approximate or preliminary estimating is to produce a forecast of the probable cost of a future project, before the building has been designed in detail and contract particulars prepared. In this way the building client is made aware of his likely financial commitments before extensive design work is undertaken.
The quantity surveyor performs an extremely important role in cost assessment, giving advice as to the probable cost of a particular design proposal and variations to it, and suggesting how similar objectives could be achieved more economically. It must, however, be emphasised at the outset that no approximate estimate can be any better than the information on which it is based. Indeed, realistic approximate estimating can be achieved only when there is full co-operation and communication between architect, quantity surveyor and building client from the inception of the scheme. It is advisable to dissuade the architect from reporting forecasts of costs to the building client until some drawings, even preliminary sketches, have been prepared and an inspection made of the site. There is a distinct possibility that the building client will endeavour to obtain an independent check on the preliminary cost figures and it is accordingly unwise to supply a high 'cover' figure. On the other hand the submission of too low a figure can lead to recriminations, as the first figure is the one that the building client will always remember. The quantity surveyor should always emphasise that an estimate based on inadequate information cannot be precise, and in such a situation he would be well advised to give a range of prices, as an indication of the lack of precision that is obtainable.
The choice of method employed will be influenced by the information and time available, the experience of the surveyor and the amount and form of the cost data available to him. It is essential to carry out a detailed site survey before a preliminary estimate is prepared as it would be quite unrealistic to assume that the site is level, and free from obstructions, has a water table well below foundation level, and that the soil has an average load-bearing capacity. Similarly, old drawings of existing buildings scheduled for adaptation need checking.
Classification of Approximate Estimating Procedures
Various precontract approximate estimating methods can be classified as follows.
Single-purpose estimates are aimed at forecasting cost and these can be further subdivided into preliminary estimates, which establish the broad financial feasibility of the project, and later stage estimates which will produce a figure comparable with that of the lowest tender.
Dual-purpose estimates are aimed at determining total costs and also the various design-cost relationships between possible variants of the projects. These estimates delve into the wider field of cost planning. The second category of estimates can be divided into two groups: primary comparative-cost estimates, which indicate the relative costs of different design solutions which will satisfy the requirements of the building client, and secondary comparative-cost estimates which apply the financial yardstick to alternatives of construction, finish and service installations applicable to the selected design.
The extent to which these various estimating processes are needed for a given project is very much dependent upon the degree of importance with which the building client views financial considerations and the size of the project. For instance, buildings let to third parties, which must show a profitable return on their capital cost and owner-occupied industrial and commercial projects where capital is limited, all need to be exhaustively and skilfully costed at each stage of development of the design.
The main problems in implementing each of the main categories of approximate estimating are now considered.
Single-purpose preliminary estimates. These are often required before any drawings are prepared and are frequently computed by rather imprecise methods. These methods include unit prices such as the comprehensive price per bed of a hospital or hotel, and an ali-in price per cubic metre of a building, whose location, shape, height, site works, services and other important characteristics may not have been determined at the time the estimate is required. These estimating methods are difficult to apply and should be used with the greatest care.
Single-purpose later stage estimates. These should only be prepared by single price-rate methods if the quantity surveyor is very experienced in the use of these methods, or if he has available as a starting point the tender particulars of a recent project which is similar in character and preferably designed by the same architect. In all other circumstances it is better to prepare estimates of this kind by way of priced approximate quantities which is a much more accurate method of estimating. In addition the total estimate can be broken down into convenient parts and justified if required, a process which is hardly feasible with single price-rate methods.
Primary comparative-cost estimates often use single price-rate methods as a basis for the computations. Nevertheless, it is important to make allowances for differing shapes and wall-to-floor ratios of alternative designs. External walling is expensive and, as a general rule, the smaller the ratio of external walling to floor area, the cheaper will be the building. The aim should be to use rectangular buildings with the sides as near equal in length as possible, provided that other aspects such as lighting and ventilation can be dealt with satisfactorily.
Similarly, it is usually more expensive to construct accommodation vertically instead of horizontally at the stage when it becomes necessary to provide a structural frame, probably at the two-storey stage in the case of a heavily loaded building, and three to five-storeys with lightly loaded ones.
Secondary comparative-cost estimates are an integral part of cost planning, in comparing the costs of possible variants in construction, finishing and servicing of the selected plan. Comparisons could be made between traditional and nontraditional construction; steel frame, reinforced concrete frame and load-bearing brickwork; various types of pitched and flat roofs; different floor, wall and ceiling finishings; and different forms of heating and lighting.
In all cases the drawing numbers on which the estimate is based should be recorded on the estimate. The date of the estimate and the allowance for price fluctuations between the estimate and tender dates should also be clearly shown. Finally, all supporting data should be filed with the estimate.
UNIT METHOD
On occasions a building client requires a preliminary estimate for a building project based on little more infol'mation than the number of persons or units of accommodation that the building is to house. The unit method of approximate estimating seeks to allocate a cost to each accommodation unit of the particular building, be it persons, seats, beds, car spaces or whatever. The total estimated cost of the proposed building is then determined by multiplying the total number of units accommodated in the building by the unit rate. Thus the mathematical process is very simple but the computation of the unit is exceedingly difficult.
The unit rate is normally obtained by a careful analysis of the unit costs of a number of fairly recently completed buildings of the same type, after making allowance for differences of cost that have arisen since the buildings were constructed and any variations in site conditions, design, form of construction, materials, etc. Variations in rates, stemming from differences in design and constructional methods, are difficult to assess and frequently there is insufficient information available to make a realistic assessment. Hence although the method has the great merit of speed of application, it suffers from the major disadvantage of lack of precision and at best can only be a rather blunt tool for establishing general guidelines, more particularly for budgetary estimating on a rolling programme covering a three to five-year period ahead. Because of the lack of precision it is advisable to express costs in ranges, with more precise costs to be determined at a later stage by more reliable estimating methods when much more detailed information is available.
The following 1975 unit rates are given as a general illustration of the application of the method to specific classes of building and are subject to all the limitations previously described.
Hospital ward accommodation: £4400 to £5000 per bed
Church hall: £130 to £160 per seat
Primary school: £460 to £540 per place
Secondary school: £850 to £950 per place
Multistorey car park: £800 to £1300 per car space
The main weaknesses of the method lie in its lack of precision, the difficulty of making allowance for a whole range offactors, from shape and size of building to constructional methods, materials, finishings and fittings, and it is not sufficiently accurate for the majority of purposes. It does, however, serve a limited number of uses such as establishing an overall target for a cost plan or calculating a sum for investment purposes, in cases where the building client or his professional advisers have considerable experience of the construction and cost of similar buildings, such as hospitals and schools. Even under these circumstances it is necessary to use this method with the greatest care and skill and with a full appreciation of its limitations.
CUBE METHOD
The cube method of approximate estimating was used quite extensively between the wars but has since been largely superseded by the superficial or floor area method. The cubic content of the building is obtained by the use of rules prescribed by the Royal Institute of British Architects6 which provide for multiplying the length, width and height (external dimensions) of each part of the building, with the volume expressed in cubic metres. The method of determining the height varies according to the type of roof and whether or not the roof space is occupied.
For a normal dwelling with an unoccupied pitched roof, the height dimension is taken from the top of the concrete foundation to a point midway between the apex of the roof and the intersection of wall and roof (half the height of the roof). If the roof space is to be occupied then the height measurement is taken three-quarters of the way up the roof slope, and with mansard roofs it is usual to measure the whole of the cubic contents.
With flat roofs, the height dimension is taken 600 mm above rooflevel except where the roofis surrounded by a parapet wall which has a height in excess of600 mm, when the height will be measured to the top of the parapet wall. If the height of the parapet wall is less than 600 mm, the minimum height of 600 mm will still be taken.
All projections such as porches, steps, bays, dormers, projecting rooflights, chimney stacks, tank compartments on flat roofs and similar features, shall be measured and added to the cubic content of the main building. On occasions it may be found that a small part of the foundations may be deeper than the remainder, and in this situation it is better to adjust the unit rate rather than to vary the cubic content of the building. Projecting eaves and cornices should be ignored when computing the volume of the building.
Where different parts of a building vary in character or function, such as a workshop with an office block frontage, then the different parts should be separately measured and priced. Basements should also be cubed separately, so that allowance can be made in the unit rate for the increased excavation and construction costs.
Features such as piling, lifts, external pavings, approach roads, external services, landscaping and similar works which bear no relation to the cubic unit of measurement, should be dealt with separately by the use of lump sum figures or approximate quantities.
The assessment of the price per cubic metre of a building calls for the exercise of careful judgement coupled with an extensive knowledge of current prices and trends. Unit prices show wide variations between different classes of building and will even vary considerably between buildings of the same type, where such factors as the proportion of walling to floor area and quality offinishings and fittings vary to a significant extent. The greater the proportion of walling in relation to the cubic contents of the building, the greater will be its cost per cubic metre.
The following typical cubic metre rates in 1975 for various building types will give a rough guide:
City bank with two floors of offices over: £45 to £70
Church hall: £25 to £40
Hotel: £60 to £80
City office block with shops on ground floor (excluding shop fronts and shop finishings, but including lifts and heating): £55 to £65
Small shop with one floor of offices over (excluding shop fronts and shop fittings): £25 to £45
Tremendous variations can occur in the cube rates ofbuildings of the same type. For instance, city offices can vary from £40 to £70 per cubic metre depending on size, shape, quality of finishings, amount of partitioning, number of fittings and a whole host of other factors, and these must all be taken into account when assessing the unit rate. With single-storey industrial buildings, wide variations in storey height can occur and costs will not vary directly in proportion to height. The roof, foundations and floor remain constant and a comparatively cheap form of cladding may be used for the walls. In this situation the cube method of approximate estimating would be illsuited and could give quite unrealistic results. A far more satisfactory approach would be to use the superficial method and to adjust the unit rate for increases in height. The cube method is, however, useful for heating and steelwork estimates.
A primary weakness of the cube method is its deceptive simplicity. It is an easy matter to calculate the volume of a building but much more difficult to assess the unit rate on account of the large number of variables which have to be considered. The cubic method fails to make allowance for plan shape, storey heights and number of storeys, which all have an important influence on cost, and cost variations arising from different constructional techniques such as alternative foundation types are difficult to incorporate in a single unit rate. Ideally, the building from which the basic cubic rate is obtained should be of similar shape, size and construction as the one under examination.
Other weaknesses are that cubic content does not give any indication to a building client of the amount of usable floor area, and it cannot readily assist the architect in his design of a building as it is difficult to forecast quickly the effect of a change in specification on the cube unit price rate. The following example may help to illustrate the approach.
SUPERFICIAL OR FLOOR AREA METHOD
In this method the total floor area of the building on all floors is measured between the internal faces of the enclosin external walls, with no deductions for internal walls, partitions, stairs, landings, lift shafts, passages, etc A unit rate is then calculated per square metre of floor area and the probable total cost of the building is obtaine by multiplying the total floor area by the calculated unit rate. Where the building varies substantially in constructiona methods or in quality of finish in different parts of the building, it will probably be advisable to separat the floor areas to enable different unit rates to be applied to the separate parts. Consideration must also be give to varying' storey heights in assessing unit rates and when extracting rates from cost analyses.
This is a popular method of approximate estimating, as it is comparatively easy to calculate the floor area of building and the costs are expressed in a way which is fairly readily understood by a building client. Furthermore most published cost data is expressed in this form. It has advantages over the cube method as the majority of item with a cost impact are related more to floor area than to volume and it is therefore easier to adjust for varying storey heights. Nevertheless, it has a number of inherent weaknesses and, in particular, it cannot directly take account of changes in plan shape or total height of the building. Similarly, difficulties are experienced in buildin up unit rates from known rates for existing buildings due to the need to make allowances for a number of variable including site conditions, constructional methods, materials, quality of finishings and number and quality o fittings.
As with the cube method, special items such as piling, heating and lift installations are normally covered by lump sums which are added to the overall cost calculated on a floor area basis. The specialist lump sums may be derived from quotations obtained from specialists or be based on information arising from previous contracts. The estimated cost of external works is usually based on priced approximate quantities.
A few typical l975 unit rates per square metre of floor area follow to indicate the range of prices, but it mus be emphasised that wide variations on these rates occur in practice.
NOTE: Costs expressed in £ per sq ft can be converted to £ per m2 by multiplying by ten and adding ten percent, for example,
£20.00 per sq ft = (20.00 x l0) + l0% = £220 per m2
STOREY-ENCLOSURE METHOD
Objectives
In 1954 James3 introduced the work of an RICS study group on a new method of single price-rate approximate estimating, termed the storey-enclosure method, with the aim of overcoming the drawbacks of the methods so far described in this article. The study group's primary objective was to devise an estimating system which, whilst leaving the type of structure and standard of finishings to be assessed in the price rate, would take the following factors into account in the measurements:
(1) shape of the building (by measuring external wall area);
(2) total floor area (by measuring area of each floor);
(3) vertical positioning of the floor areas in the building (by using greater multiplying factors for higher floors and greater measurement product for suspended floors than nonsuspended);
(4) storey heights ofbuilding (proportion of floor and roof areas to external walls);
(5) extra cost of sinking usable floor area below ground level (by using increased multiplying factors for work below ground level).
Nevertheless, the following works would have to be estimated separately:
(1) site works, such as roads, paths, drainage, service mains and other works outside the building (these are best covered by approximate quantities);
(2) extra cost of foundations, which are more expensive than those normally provided for the particular type of building (again this is best covered by approximate quantities);
(3) sanitary plumbing, water services, heating, electrical and gas services and lifts (priced approximate quantities or price from specialist consultant);
(4) features which are not general to the structure as a whole, such as dormers, canopies and boiler flues (separate priced additions);
(5) curved work.
The storey-enclosure method consists basically of measuring the area of the external walls, floor and ceiling which encloses each storey of the building. These measurements are adjusted in accordance with the following set of rules.
(1) To allow for the cost of normal foundations, the ground floor area (measured in square metres between external walls) is multiplied by a weighting factor of two.
(2) To provide for the extra cost of upper floors, an additional weighting factor is applied to the area of each floor above the lowest. Thus the additional weighting factor for the first suspended floor is 0.15, for the second 0.30, for the third 0.45 and so on.
(3) To cover the extra cost of work below ground level a further weighting factor of one is applied to the approximate wall and floor areas that adjoin the earth surface.
Summing up, the procedure is to take twice the area of the ground floor, if above finished ground level, and three times the area if below (measured between external walls). It will take once the area of the roof measured on plan to the external face of the walls (the same area whether a flat or pitched roof), twice the area of upper floors (to cover work above and below, including partitions), plus the appropriate positional factor; it will also take once the area of external walls above ground measured on their external faces (ground level to eaves), and twice the area below ground in basements, etc.
All the areas are multiplied by the appropriate weightings to obtain the storey-enclosure units which are then totalled. To obtain the estimated cost of the building, the total of storey-enclosure units in square metres is multiplied by a single price-rate built up from the costs of previous similar projects. The cost of external works and other special items can be added. Basically, the aim of this method is to obtain a total superficial area in square metres to which a single price-rate can be attached, and the effect of the various rules that have been outlined is to apply a weighted cost factor to each of the main parts or elements of the building.