Konut Projelerinde, Ön Tasar Aşamasında, Maliyet Tahmini İçin Bir Model

Abstract

Bu çalışmada, maliyetin büyük bir kısmının ilk tasar aşamalarında oluş tuğu düşünülerek, ön karar aşamasında kullanılabilecek, tasarım değişkenleri ile maliyet arasındaki ilişkiye dayanan bir maliyet tahmin modeli geliştirilmeye çalışılmıştır. Giriş bölümünde bu ana amaç ve konunun işlenişi belirlendikten sonra, ikinci bölümde, tasarım aşamasında nitelik kararlan henüz alınmadığından, maliye ti etkileyen niceliksel tasarım değişkenleri yerleşme, bina ve eleman ölçeğinde incelenmiştir. Bina ölçeğindeki niceliksel tasarım değişkenlerinin, yapı alt sistemlerini nasıl etkilediği de bu bölümde araştırılmıştır. Üçüncü bölümde, maliyet tahmininde kullanılan modeller genel olarak incelenmiştir. Model seçimini etkileyen faktörlerde gözönüne alınarak, tasarım değişkenleri ile maliyet arasındaki ilişkiden yola çıkıldığında maliyet hesaplamaya en uygun modelin hangisi olabileceği saptanmaya çalışılmıştır. Dördüncü bölümde, bir önceki bölümde ön tasar aşamasında kullanılması uygun olan modellerden biri olduğuna karar verilen, regresyon analizi tekniğinin tanımı ve kullanımına yer verilmiştir. Beşinci bölümde ise, ikinci bölümde anlatılan bina ölçeğinde maliyeti etkileyen tasarım değişkenlerinden elde edilen göstergeler ile maliyet arasındaki iliş ki, regresyon analizi ile saptanmış, bu ilişkiye dayanılarak konut projelerinde, ön tasar aşamasında, maliyet tahminine yönelik bir model elde edilmiştir.

A great part of building cost is formed from the first decision processes. As the decisions made during the first decision processes will form inputs for the next process, cost in use is actually a result of first investment decisions. Therefore cost estimation has a great importance in the pre-design process, during which the data one has are usually about the quantitative design variables. Regarding this as a starting point, the aim of this thesis is to develop a model for cost estimation depending upon the relation between the design variables and the cost. Therefore, design variables and then cost estimation models are examined and due to the relation between the variables, the model of regression is agreed to be the most appropriate estimation model to use. In the second chapter, design variables and their effects on cost is defined. Quantitative design variables are analyzed in site, building and element scale. I. COST IMPLICATIONS OF THE DESIGN VARIABLES IN SITE SCALE The most important design variables in site scale are related to the usage of the building site, which has an important effect on first investment cost and cost in use. Also appropriate usage of the site decreases the sub construction cost. II. COST IMPLICATONS OF THE DESIGN VARIABLES IN BUILDING SCALE The most important design variables in building scale are plan shape, storey height, number of storey, external envelope area and circulation space. a) Plan shape: The shape of a building has an important effect on cost. As a general rule, the simpler the shape of the building, the lower the unit cost will be. The full effect of plan shape is not always easy to establish, but a simple method can be devised for relating the amount of external walls to the superficial floor area of the building so as to reveal the effect of different plan shapes vi b) Storey heihgts: The main constructional items which will be affected by a variation in storey heihgts are walls and partitions together with their associated finishing and decorations. Some subsidary items such as heat source, pipes, cables, staircases, lifts, etc. might also be affected by an increase in storey height. c) Number of storey: Tall buildings are invariably more expensive to build than two or three storey buildings. The only exception to this rule is that the addition of one or two storeys to the design of a tall building for the purpose of making the best use of the lifts or other expensive services, may slightly decrease the cost per storey. d) External envelope area: The external envelope of a building, that is the external walls and roof which enclose it, is an important factor in the cost of a building. A square building is inherently economical in wall area, but the total envelope cost will depend upon the number of storeys. e) Circulation area: Circulation areas in entrance halls, passages, corridors, stairways and lift wells, can all be regarded as "dead space" which cannot be used for a profitable purpose and yet involves considerable cost in heating, lightening, cleaning, decorating and in other ways. IH. COST IMPLICATIONS OF DESIGN VARIABLES IN ELEMENT SCALE Cost implications of desing variables in element scale can be classified in three groups, as follows: a) The shape and construction of the elements b) The size and the weight of the elements c) Amount of the production Changing the design variables in building scale affects the building sub-systems which constitute the cost of the building. Different groupings can be made to analyze a building from the point of its sub-systems. In this thesis, it is as follows: 1. Substructure and foundations: They are affected by number of storeys. 2. External partitions: They are affected by plan shape, storey heights and number of storey 3. Internal partitions: They are affected by plan shape, storey heights and number of storey. vu 4. Horizantal partitions and staircases: They are effected by plan shape and circulation area. Staircases are affected by number of storey and storey heights. 5. Roof: It is affected by plan sphape and number of storey. 6. Windows: They are affected by external partitions. 7. Internal and external doors: These are controlled by the same factors as internal and external partitions. 8. Frame: It is affected by planshape and number of storey. Cost models used at design sketch: Cost is one of the measures of function and performance of a building and should therefore capable of being modelled in orter that a design can be evaluated. Cost modelling may be defined as the symbolic representation of a system. Objectives of modelling can be listed as follows; 1. To give confidence to the client with regard to the expected cost of his project. 2. To allow the quick development of a representation of the building in such a way that its cost can be tested and analysed. 3. To establish a suitable system for advising the designer on cost that is compatible with his own build up of the design. 4. To establish a link between the cost control of desing and the manner in which cost are generated and controlled on site. Around this ejective requirements of a good cost model should incorporate the following criteria; 1. The data requirement for the model should be freely available in the form and amount available. 2. The model should allow for continous updating by incorporating new data that become available. 3. The model should be capable of evolving to suit the needs of a changing stiuation that prevalent in the construction industry. 4. The entire process of cost model management should be able to be done quickly, cheaply and efficiently. 5. The model should accurately and reliably represent that which it is attempting to predict. A very difficult subject about cost models is assesment of the model performance. The factors which affect model performance are as follows; -Data - Data/model interface - Model technique vui - Interpretation of output. Cost models can be classified in different ways. One of them is the following classification; I. Traditional cost models (in-place materials methods) a) Single price estimating models - The unit method. - The cube method - The superficial area method. - The storey enclosure method. b) Elemantal estimating c) Operational estimating d) Resource related methods II. Contemporary Cost Models a) Causal or empirical models. b) Regression models c) Simulation models. Choice of the appropriate model depends on the following factors, - purpose of the cost prediction - amount and quality of the data which is used to predict the cost. - characteristic of the work which is predicted. - cost of the modelling. - Time dedicated to the cost estimation. As a result, regarding the aim of this study together with the factors affecting the choice of a model, regression models are agreed to be one of the most appropriate models for this studs. Mathematical functions or equations express the exact relationship present among the variables of interest. If we substitude numerical values for the variables on the right-hand sides of these equations, exact values of the quantities on left - hand sides can be calculated. Especially in the social scence exact relationships are not generally observed among variables, but rather statistical relationships prevail. That is certain average relationships may be observed among variables, but these average relationships do not provide a basis for perfect predictions. The term "regression analysis" refers to the methods by which estimates are made of the values of a variable from a knowledge of the values of one or more other variables, and to the measurements of the erros involved in this estimation process. The term "correlation analysis" refers to methods for measuring the strength of the association (correlation) among those variables. IX The first purpose of regression analysis is "to provide estimates of values of the dependent variable from values of the independent variable. The second purpose of regression analysis is to obtain measures of the error involved in using the regression line as a basis of estimation. The third objective, which is classified as correlation analysis, is to obtain a measure of the degree of association or correlation between the two variables. A regression model is a method of determining the relationship between variables by the "method of least squares" which seeks to minimize the sum of the squares of the difference between the observed values and the predicted vaules. In two variable lineer regression, the expression for the straihgt line is of the form; y=a+bxa=y-bx, b= Where "a" is the intercept of the line with the y axis and "b" is the slope of the line. In multiple or non-lineer regression analysis, the equation can be written as follows, y=a+b1x1+b2x2+..+bnxn a°d logy « a + b\logxx + b2İogx2 +... + bnlogx" or y=a+bx+cx2 Correlation between two variables can be defined by determination coefficent, correlation coefficent and t test. As a result, design variables have been examined from the point of view of the aim of this study, and a model has been developped by establishing the relationship between cost and the indicatiors consisting these variables. For this purpose 30 housing projects are analysed and their costs are estimated by the use of a computer aided cost estimation programme, named as MBB. When 1 square meters of construction cost is taken as the independent variable, while 1 square meters of profitable space cost is being taken as the dependent variable in the indicators consisting of desing variables, the equations are as follows, BM-584 226+232 840 DJCA-BİA+1 343 688 TA-BİA + 414 239 FA-BİA-1742 FA-BKS FABM» 2323876 + 275455 DKA-BİA + 1961 234 TA-BİA - 1753291 FAB-BtA But, it must also be taken into account that with these equations, the cost will be estimated approximately in the frame of some acceptances.