Neural network approach to grading maintainability of wet areas in high rise buildings

1.3 RESEARCH PROBLEM The level of maintainability of a building is proportional to its life cycle cost expenditure, to keep it in optimum performance.. If the building is not maintainab

CHAPTER 1 INTRODUCTION

1.1 BACKGROUND

The importance of recognizing and assessing the total costs of owning, constructing, operating and managing buildings is being increasingly realized throughout the developed world [1] Quality costs in the construction industry as a whole are relatively high in terms

of total project costs The bulk of this project costs are attributed to the building’s operation and maintenance costs during its lifespan [2-3]

It is imperative to understand the degradation processes of buildings, to define methods and tools in order to easily identify potential defects (at design stage, as well as in servicing stage) and to find solutions to reduce their effects and to avoid any unforeseen cost for repair or maintenance Premature failures cause an enormous waste of resources [4] The life cycle cost of buildings, as a matter of fact, has a big weight over gross world product [4]:

o Building activities represent approximately 40% of the world production

o Solid waste production from building construction, demolition and the production

of building products approximately represent 40% of the world waste production

Moreover, building failures approximately generate 10% of that life cycle cost [4] Thus failure prevention could have very substantial effects both on the direct costs of a building and the environmental impact of construction sector, in every country It is of the utmost importance to design for durability and maintainability of buildings and their products With this in mind, it should be of primary interest to adopt a life cycle cost (LCC) perspective to building maintenance management

Buildings that are managed with rational and long-term perspective will remain attractive for a longer time period and the need for replacement will be lessened As buildings are getting more technically complex with an increasing number of installations and equipment, these installations usually have shorter life spans than the building itself It is suspected that this would increase maintenance costs compared to older buildings due to the accelerated aging of components and installations This implies that components might have to be replaced even before the end of their technological lifespan This replacement can contribute to a sharp increase in operation and maintenance costs which in turn increases the total building life cycle costs

Coupled by the lack of structured objective condition rating system, and means of analyzing and reporting key building information, it is impossible to simultaneously assess current conditions accurately, project future conditions, and track building performance [5] or even building costs Therefore, key building components cannot be properly evaluated, nor can deficiencies be identified When defining the maintenance

This dissertation would provide the impetus to spearhead serious considerations and create

a momentum for the Singapore construction industry to incorporate maintainability issues into the design stage

project, from design to construction to maintenance, should be seen in totality It recognized that the higher costs of maintenance of such buildings is often hidden and not considered when assessing the cost of the construction project Benchmarks that can be used to audit maintenance costs and produce manuals which give the design life and maintenance costs of components were deem to be desirable [15]

Wet areas is one that poses great threats to the maintenance budget of buildings The annual maintenance expenditure for residential and office buildings in Singapore accounts for $35/m2 and $87/m2 respectively [16] Defects in wet areas such as toilets and bathrooms have caused inconvenience to the users and are aesthetically unpleasant

Compounded by poor construction workmanship to provide watertightness, difficulties in maintenance have contributed to increasing number of occurrence of defects in wet areas Maintenance managers are becoming painfully aware of the lack of knowledge and tools

in their organizations to assist in effectively carrying out any planned maintenance The management could only react to anomalies, which are the pre-acknowledgement of defects, when the defects became obvious

In order to accomplish building maintenance in such a scenario, ad hoc maintenance seems to be the next best approach Consequently, the funds needed to repair even the simplest defect would increase several folds [5] Further, when major components eventually fail, repair work may be deferred due to budget constraints In the short term,

The evaluation of building maintainability should be a methodological approach whereby the different measurements of defined variables are analysed to trace the causal relations between the measures and the specific defects There should be in place a maintainability assessment model for the evaluation of whole life performance; with regards to the important parameters of determining the ease of maintenance, right from the start of the project development stage

There are many reasons why it is important to develop a predictive maintainability model:

1 to spearhead the incorporation of maintenance issues into consideration during design stage,

2 to establish a benchmark by which the maintainability of wet areas can be measured and compared,

3 to serve as a tool for the prediction of maintenance costs related to wet areas over the building’s lifespan

1.3 RESEARCH PROBLEM

The level of maintainability of a building is proportional to its life cycle cost expenditure,

to keep it in optimum performance If the building is not maintainable and defects are recurring, more money has to be spent to restore buildings to near original conditions which consequently impose more cost burdens to the total life cycle cost of building With large amount spent on the maintenance of buildings each year, prudent examination of influencing factors of design, construction and maintenance is necessary [9, 25-28]

With merits of a maintainability model for wet areas highlighted above, it is necessary to develop a modeling technique to improve design and construction practices in the building industry with an aim to prolong the whole life performance of wet areas and minimize maintenance costs Further reasons to push ahead with the construction of maintainability model include the following

First, though wet areas do not constitute a large area of a building, the maintenance costs spent on wet areas alone can be as high as 50% of a building’s maintenance costs [29] Thus it is essential to identify the important parameters that would influence the aesthetic and functional performance of wet areas

Secondly, very few attempts have been made on improving maintainability of buildings

by considering maintenance issues at the design stage If such issues can be thoroughly dealt with during the inceptual stage, decisions made during design stage would have greatest effects to prolong the performance in wet areas

Thirdly, the model would assist in deducing the important design parameters to derive optimum design layout that would facilitate maintenance within the scarce maintenance budget The use of a more superior neural network model is to ensure all possible underlying non-linear patterns are captured This derived model would aid designers by giving more weight to the various parameters that would influence the future level of maintainability of wet areas

The objectives of the study are:

1 To determine common defects and their sources;

“Prediction of level of maintainability of wet areas can be facilitated by the use of neural

network based model”

1.6 SCOPE OF WORK

1 A thorough literature review of relevant published books, past dissertations, journals, research digest and other related articles on the issues of the maintenance issues of wet areas was conducted Through this review, a better understanding on the theoretical aspect of the study and current status of the maintenance field in the industry can be achieved

2 Condition surveys were conducted to determine the types and extent of the common defects in different building types and their causes Types of

defects from the relevant establishments, building managers, repair contractors and case studies were also collated to determine the common defects in non-residential buildings

3 Analysis of defects and identification of factors affecting maintainability of wet areas

4 Development and validation of the model

1.7 ORGANISATION OF THESIS (Figure 1.2)

Chapter one gives brief background on the importance of building maintenance in the construction value chain In the local industry one of the building elements, which effects high maintenance costs, is the wet areas Thus the issue of enhancing maintainability of

Analysis of defects

Construction of inputs to model

Figure 1.1 Scope of works

Construction of predictive model Validation.

wet areas through modeling maintainability using neural network was emphasized in this chapter, which takes the form of a statement of the research problem A concise illustration of the research objectives and work scope would also be included

A conceptual review of issues of building lifecycle, building performance and building maintenance which lead to the notion of improving building maintainability was highlighted in Chapter two A review of risk management was also included to set the foundation for use as the approach to this research An evaluation of two commonly used forecasting techniques of multiple linear regression and artificial neural network seeks to justify the choice of using the more superior neural network to model wet area maintainability

Chapter three presented the research methodology This served to guide the conduct of this research and a concise flow of the development of the thesis

Chapter four presented common defects in wet areas as assessed from the building samples Photos and self-drawn diagrams would be provided to facilitate the comprehension of underlying mechanisms of the defects This would provide sufficient evidence to substantiate the development of the various influencing factors of maintainability

Chapter five furnishes a comprehensive review of the influencing factors identified to have far reaching effects on the maintainability of wet areas, namely design, construction, material selection, maintenance, microenvironment and building profile Detailed technical knowledge is drawn upon to provide factual information on how these parameters would directly or indirectly affect the occurrence of defects

Chapter six sets out the guidelines for the implementation of the neural network model The quantification of identified parameters would present to give proper guidelines as to how this research grades each variable These details were researched based on a wide literature review, surveys and interviews from various industry participants, giving as accurate as possible the ranking criteria and desired technical details

Chapter seven brings us to the core of the research with the development of the maintainability model - WETAMS Results were displayed to illustrate the predictive power of the model The more influencing factors would be identified via a sensitivity test analysis

Chapter eight gives conclusions and highlights important findings of the research Further recommendations were included to give directions to future work in building maintenance

o Brief background

o Justification for research

o Research objectives and scope of work

Chapter 4: Defects Analysis

o Common types of defects

o Sources of defects.

Chapter 2: Literature Review

o Review of building performance, building life cycle cost, building maintenance, risk management.

o Issue of maintainability in Singapore

o Selection of neural network for use in research

Chapter 5: Factors affecting

maintainability of wet areas

o Design, construction, maintenance

o Ranking criteria of factors

Chapter 7 Results and Discussions

o Illustrates the predictive power of model

o Sensitivity test of model to examine more

important factors

Chapter 8 Conclusions and Recommendations

o Concludes the whole study with significant findings

o Gives further recommendations

Figure 1.2 Organization of thesis

1.8 MAJOR CONTRIBUTION OF THESIS

The persistent occurrence of repetitive defects in wet areas demands a renewed comprehensive analysis of the various parameters influencing the performance of areas over the building lifespan

In this thesis, the first attempt to construct a novel maintainability model, using artificial neural network, to spearhead continued improvement in area of building maintenance is presented This first initiative deals with internal wet areas The thesis describes how the neural network can be used to predict the level of maintainability and estimate the importance of the various parameters that would influence the ease of maintainability of wet areas in terms of real cost values It provides an effective alternative to guide the designers in focusing on details that would have greater impact on the performance of the wet areas as a whole system; thus enabling reduction of resource wastages and low productivity-value maintenance work right from the design stage

This chapter seeks to establish a conceptual review on issues of building maintainability with respect to building performance of wet areas and the building life cycle cost involved

in maintaining them A brief introduction to the concept of building pathology is also provided A brief review of risk management is included as the approach in evaluating maintainability, where the subject in view is “achieving high level of maintainability” and the various risk sources are design, construction, maintenance, microenvironment and building profile

The notion of maintainability has been manifesting during the past decades as building owners demand more durable buildings The issue of maintainability thus has grown to paramount importance, as owners are confronted with soaring maintenance costs [2, 6 28, 34-39] These lofty maintenance costs can be reflected in the increasing number of defects attributed to deficiencies of design [14, 27], substandard workmanship [9] and incompetent maintenance practices [13]

This trend is due to the growing complexity of buildings, such as increasing proportion of systems in them, higher levels of service requirements, and the higher portion of maintenance costs added to the life cycle costs of buildings Thus to address the issue of maintainability in totality, the marriage of building performance and building life cycle cost is of importance

The term ‘maintainability’ of buildings may then be expressed as ‘achieving the optimum performance throughout the building’s life span within a minimum life cycle cost’ (Chew

et al., 2003) (Figure 2.1) This can be achieved through an optimum mix of design,

construction, material and maintenance attributes This principle can be simply expressed

where Μ is a set of scalars providing points to an underlying maintainability function

k= the expected performance

)(Risk

∫ = mix of risk attributes

Maximize performance

Minimize cost Minimize risk

Maintainabiiy

Figure 2.1: Function of maintainability

2.3 MANAGEMENT OF RISK IN MAINTAINABILITY

Grading of the maintainability requires the exploration of the underlying risk environment

of buildings’ LCC, expressed in the following section, focusing on all the important risk factors of maintainability [4, 40] A systematic approach to risk management consists of four major steps: risk identification, risk analysis, risk monitoring and a response process (Figure 2.2) A cause and effect approach is adopted here to investigate, analyze and identify the more important factors, hence, forecast maintainability with these factors It is certainly desirable to have a model to evaluate the level of maintainability via “grading” system that has complied with particular specifications to evaluate and predict the risk

According to Raffery (1994), risk and uncertainty are simply defined as the deviation of actual outcome from its forecast value [41] Al-Bahar and Crandall (1990) extended this concept further by defining the same as “an exposure to chance of occurrence of events adversely or favourably affecting the project objectives as a consequence of uncertainty” [42]

The origin of risk is the uncertainty inherent to any subject, and every risk is associated with at least a cause, a consequence at least if it occurs, and the probability or likelihood

of the event occurring There are known risks that can be analyzed and managed, and other unknown risks that can be addressed [43]

2.3.1 Risk Identification

The very first step in the risk management process is to have clear identification of risk factors and their sources applicable for the subject in view Once the risk is identified and defined, it becomes a management problem [44]

2.3.2 Risk Analysis

Once the significant factors affecting maintainability have been identified, they can be analyzed in depth Such risk analysis is concerned with the severity of each type and source of risk on maintainability

2.3.3 Risk Response

At the end of risk phase, the next task would be to decide how to response

to such risk According to the construction industry practice, there are four major risk response methods known as risk avoidance, risk reduction, risk transfer and risk retention [44-45] To deal with maintainability, the more rational approaches include risk avoidance and risk reduction This can be carried out by making effective maintenance-enhancing decisions during the design stage

2.3.4 Risk Monitoring

Risk monitoring should be conducted after the first three steps were completed This takes place when effective building management is to be carried out Reviewing and assessing the risk management results throughout the lifespan of the building is an effective way to improve the risk management procedures

2.4 BUILDING DEFECTS

Building defects were defined as results of failures or shortcomings in the function, performance, statutory or user requirements of the structure, fabric, services or other facilities [67] They hinder optimum building performance and impose high costs to scarce maintenance budgets Many defects have escalated with the growing demand in buildings The fragmented nature of building and construction process is found to be the

main contributing factor [7, 68-73] However, building defects may be an issue only if they are major or excessive [73]

To effectively improve the ease of maintainability, it is of a compelling logic to examine

in great details the bane to maintainability -persistent repetitive occurrence of building defects All buildings eventually develop defects when exposed to the wear and tear [67] Proper maintenance and the avoidance of human error may delay the process However in order to avoid or minimize building defects, it is imperative to ensure that minimum hiccups or deficiencies arise during the entire course of construction

Researchers worldwide have highlighted the importance of awareness of the sources of defects [8-9, 29, 72, 74-78] Main causes of building defects can be classified into a few areas [25-26, 79-80]: design detailing, materials selection, construction quality, microenvironment and maintenance practices (Figure 2.3)

2.4.1 Design

Fifty eight percent of defects were found to originate from faulty designs in

a survey conducted by the Building Research Establishment [81] Faulty designs are found to contribute to an increase in cost of maintenance [12-

13, 82] Many problems arise when design is satisfactory in principle but has low probability of achievement in practice [83] Common design inefficiencies include:

Failure to follow well established design criteria in the choice of structural system and selection of materials

Ignorance of the basic physical properties of the materials

Use of new materials or innovative forms of construction which have not been properly tested for use

Misjudgment of climatic conditions under which the material has to perform

Poor communication between design and construction teams

performance requirement of the materials life of the buildings

methods of cleaning

nature of environment

relationship with other materials

aesthetic requirement

clients’ functional needs

meet budget constraints laid down

can be maintained in a good working order for reasonable period of time and at

a reasonable cost

2.4.3 Construction Quality

Having the best design practices but built or constructed erroneously would generate more defects Poor construction practices such as inadequate concrete

curing and incorrect application of waterproofing membrane manifest themselves into greater problems when there is a lack of proper supervision and inspection of works [9] The lack of training and skill development affects the operational competency of operatives such that workmanship quality is reduced and resulting in more construction errors and defects [76]

2.4.4 Microenvironment

Microenvironment includes the surrounding conditions such as usage and ventilation In the tropics, high humidity poses dampness problems This trapped dampness in materials, together with oxidation and thermal effects would consequently lead to defects occurrence of different nature and extent

Maintenance practices and management involves the effective deployment of all resources in the servicing, rectification and replacement of different defective components [80] Essentially servicing refers to the routine maintenance operations undertaken to upkeep the performance of building components Rectifications are performed when shortcomings in design and construction of components transformed into defects in the early life of the buildings When such defects worsened till total failure is inevitable, replacement is the next alternative However if maintenance practices are not appropriate and could not make good any defects, this could be another source of inadequacy Regular cleaning and servicing schedules, together with a system of good inspection and fault reporting for early repairs would reduce the need to go for costly repairs and even replacement [84-85]

2.5 DEFECTS IN WET AREAS

2.5.1 Cracks

Concrete Society Technical Report No.22 (1992) on non-structural cracks in concrete explained the principals, which govern the formation of cracks in concrete and defined the various types of non-structural cracks which might occur Figure 2.4 shows a comprehensive “family tree” of crack types Cracks are mainly due to the intrinsic nature of the concrete and its constituent materials This report also highlighted the 3 main types of intrinsic cracks [86]:

(a) Plastic cracks which appears in the first few hours;

(b) Early thermal contraction cracks which appears at a time between

one day and 2-3 weeks;

(c) Long-term drying shrinkage cracks which appears thereafter

2.5.2 Paint Defects

An important factor for satisfactory painting is the dryness of the wall Excessive moisture affects the ability to apply most type of paint and affects the long-term performance It can cause the deterioration of many materials and the movement of salts and promote the growth of moulds

Hinks and Cook (1997) presented the common causes of paint defects are [87]:

(a) Damp background, high porous background, dirty background;

(b) Dimensionally or chemically unstable background;

(c) Incompatibility between background and paint;

(d) Permeability of coating;

(e) Loss of adhesion; and (f) Poor paint quality, uneven application

“Good Industry Practice Guide Painting (2001)” published by BCA stated the common types of paint defects during service life of a building This includes efflorescence, deterioration/ erosion of pigment, yellowing, saponification, chalking, peeling and flaking paint, blistering, staining, rust stains and algae and fungus growth [88]

2.5.3 Water leakage

Chew and De Silva has expressed in two papers that water leakage is one of the most common defect in wet area [29,71] The factors affecting watertightness include improper application of waterproofing at penetrations and incorrect fixing

of discharge pipes including lack of proper gradient for water flow towards the drainage outlet and the “short” gully pipes or the “short” discharge pipes

Plastic Shrinkage Plastic Settlement

Shrinkable Aggregates Drying Shrinkage Crazing

Corrosion of Reinforcement Alkali-Aggregate Reactions Cement Carbonation

Freeze/ Thaw Cycles

External Seasonal Temperature Variations

Early Thermal Contraction Accidental Overload Creep

2.5.4 Fungi and Algae Growth

Fungi and algae growth is a common defect found in wet areas under humid tropical conditions These organisms thrive where there is ample supply of moisture, nutrients and sunlight The main moisture sources found are water retention from improper pipe connections, air-con ducts and external walls Further, some of these construction and finishing materials are highly susceptible to bio-deterioration

2.5.5 Cracking and Spalling of Concrete

Cracking and spalling of concrete due to carbonation is found to be a significant problem resulting from long-term water retention in concrete Porous, permeable concrete and insufficient concrete cover thickness are testimony to rapid carbonation Low quality concrete may carbonate to a depth of 25mm in less than

10 years [7, 68-69]

Under these conditions, there is a high tendency for the reinforcement to corrode

In the corrosion process, the transformation of iron to rust is accompanied by an

increase in volume This increase in volume leads to form cracks Further, with

insufficient concrete cover around the reinforcement, spalling results

Cracking of tiles belongs to the next category of defects It is established that these defects usually occurs just a few months after building completion Shrinkage and high porosity of concrete attributed by poor mixing of screed or tile bed leads to the formation of these Debonding of waterproofing layer has also been found as a source of cracking and crazing of tiles [29]

2.5.7 Pipe leakage

Pipe leakage arises when pipe system has deteriorated Cast iron were common

material used in the past Over time, when wastewater is carried along these pipes,

there is a likelihood that corrosion of metal will occur

Very few research works were done specifically for wet areas Most reviews were contributed to the general description of defects due to poor design detailing, construction, material and maintenance with no relation to the entire system of wet areas

The author seeks to bring in a new perspective of the study of defects where defects in wet areas are usually brought about by a complex combination of subsequent chemical reactions, fair deterioration, poor design and construction, inappropriate selection of materials and maintenance The mechanism of how defects were formed was illustrated in diagrams to give visual impact to readers in perspective of wet area system

Building pathology draws attention to [90]:

1) identification, investigation and diagnosis of defects in existing buildings

2) prognosis of defects diagnosed, and recommendations for the most appropriate course of action having regard to the building, its future and resources available; and

3) design, specification, implementation and supervision of appropriate programmes of remedial works, and monitoring and evaluation of remedial works in terms of their functional, technical and economic performance in use Understanding the principle and the practical application of building pathology could assist all interested building parties to achieve total building maintainability

2.7 MAINTAINABILITY MANAGEMENT

Improving maintainability not only requires efficient building maintenance, but also demands a concentrated effort from the design right through to the construction stage There are many factors in the design and construction process that can determine whether

a structure would meet its design life There are examples of problems that have occurred where structures have been exposed to aggressive environments and where design and workmanship have been poor [91] Problems have also arisen as a consequence of a design philosophy in which minimizing initial costs was dominant [22] Premature and costly repair or even replacement may be necessary Thus to attain the optimum ease of maintenance, it is not easy to balance the demand of maximum building performance and the need to minimize building life cycle costs

Building maintainability is a complex issue which involves decision-making in optimizing the selection of building elements In general, this is not a straightforward task given (1) the large number of materials and variety of designs (2) variability of the performance (3) the myriad of latent risk of failures associated with different systems and components

Extensive researches have been conducted to improve the maintainability of buildings in the past These proposed benchmarks include the establishment of material durability guidelines, life cycle costing database and identification of attributes in preventive maintenance programmes, evaluation of their effectiveness and prioritization, maintenance budget allocation and management [3, 11, 92-94]

2.8 REVIEW OF PAST WORKS

There are in place various initiatives and benchmarking tools available overseas The following highlights some systems to improve the quality of building and subsequently reduce maintenance needs

In USA, several organizations including the U.S Navy and the Army uses systematic inspection systems called Building Engineered Management System (BUILDER) to assess the condition of buildings [5] In this system, indexes ranged from 0 – 100 are divided into seven condition intervals such as 0-10: failed, 10-25: very poor, 25-40: poor, 40-55: fair, 55-70: good, 70-85: very good, and 85-100: excellent, to provide the quantitative means for the observable defects The Building Owners and Managers

Association (BOMA) of United States has also compiled a collection of income and expenditure data analyses ranging national cross tabulations and social building data tabulations to city analyses for 128 cities in North America This is to establish a benchmark for maintenance costs The data is collected via comprehensive survey form For each metropolitan area, the data is broken down into the location and size of development The statistics are published annually in an Experience Exchange Report (EER) for industry use This EER serves as a mean by which building owners and managers can measure their own performance against an industry benchmark and also as a way by which trends in maintenance income and expenditure may be deliberated

In the Netherlands, the construction industry has been inclined towards the development

of techniques in order to extend service life of existing structures, rather than drawing up measures to ensure maintainability of buildings A Condition Assessment Score is formulated [16] as a means to prioritize maintenance expenditure A Condition score of 1

to 6 is assigned to building components after systematic investigations Scores can be used

to illustrate the effect of executed maintenance and as a framework for interpreting maintenance costs by computing the cost required to improve the score

Lam (2000) commented that building performance could affect the quality of commercial buildings A survey conducted in Hong Kong in June 2000 revealed that up to 40% of maintenance faults were related to design, up to 30% related to construction or installation; and up to 30% related to maintenance management [95] The application of Quality Management and Assurance System via performance audit, can serve as useful framework for parties involved in design, installation and management of buildings

Similarly, in Belgium, a quality management approach using “quality system standards” stipulated in BS 5750/ISO 9000 was formulated All elements in the process of design, fabrication and field performance were to follow an integrated approach [96] A quality chain and tracking system (clients, designers, contractors, manufacturers and maintenance specialist/provider) should also be put in place to check linkage developed during design,

construction, and between user requirements, to verification of the completed building performance [97-98]

Building Quality Assessment (BQA) is yet another quality control tool used in New Zealand [99-101] The procedure in BQA prescribed audit of office building quality and key areas for assessment Nine main categories that make up BQA profile are

“presentation, space, access and circulation, business services, personnel amenities, working environment, health and safety, structural considerations and manageability” With complete compliance with this assessment, minimum building maintenance would

be expected

The Housing Association Property Manual (HAPM) Technical Audit Unit (1997) undertakes a more thorough approach to ensure the quality design and construction Three performance comparisons, viz., Design Quality Indicator (DQI), Workmanship Quality

Indicator (WQI) and the Maintenance Indicator (MI), which consists of individual Component Maintenance Indicator (CMI) and Whole Scheme Maintenance Indicator (SMI), were provided The CMI and SMI use data from Component Life Manual, where information on the expected durability of numerous building components are listed Each

of the indicators is ranked on a scale of 0-100, thus with the achievement of a higher score, the better the quality of each stage is registered This would follow that if the construction product is designed and constructed soundly, minimum maintenance would

be required [102]

as guides to future resolution of similar defects

“Performance based specification system” to facilitate owner in considering design alternatives to achieve desired maintenance objectives and thus have lesser impact on LCC was proposed [103] Owner identifies performance objectives focusing on:

• customer needs, technical system and procedures, continuous improvement, employee involvement and integrates with maintenance strategy /business goals

• Designers consider alternative solutions, expected failures and remedial actions

• Performance data collected for possible design solutions

• LCC model compares design alternatives and select best system proposal

From the above, it has been shown that some countries already practice total building performance audit, benchmarking and quality management procedures and life cycle costing yardsticks in guiding developments towards improved maintainability The review of these existing systems only reveals that building maintenance has been a major

Construction 21 Steering Committee (CTC) set up by the Ministry of Manpower (MOM) and Ministry of National Development (MND) in 1999 has identified in Strategic Thrust 3

“the enhancement of maintainability of the buildings” as one of the areas of focus, which would help attain greater productivity and other breakthroughs To achieve greater

efficiency during the whole lifecycle of the project, it is indisputable to view all project stages in totality The Committee strongly proposed for the establishment of a maintainability scoring system to grade buildings accordingly, as a major step to upgrade the Singapore construction industry [15]

2.9.1 Framework for Improving Maintainability

There is a growing need for more effective handling, more ready access to available knowledge and feedback To better manage and enhance the quality of maintenance management, an awareness on issues influencing maintainability and benchmarking the level of maintainability are important fundamentals which are indispensable This framework includes creating a momentum for the industry to learn about the problems and obstacles to building maintenance and the development of a benchmark to aid in the incorporation of maintenance issues in upstream construction

2.9.1.1 Improving the knowledge

Faced with difficulties in maintenance with increasing number of occurrence of defects, maintenance managers are becoming painfully aware of the lack of knowledge and tools in their organizations On the other hand, designers are not well focused on achieving maintainable designs, owing to poor awareness of the maintainability problems during the service life of building The study was set to collate existing problems of maintainability issues pertaining to wet areas to create

an awareness of the current obstacles to good maintainability under tropical conditions

2.9.1.2 Setting Maintainability Benchmarks

It is of convincing logic to improve the level of maintainability right from the design stage This can be done by an objective measuring tool The rational is to enable the decision maker to choose a more maintainable design option No previous attempts were made to develop a scoring system to grade maintainability

In this regard, this research would be the very first one to grade maintainability of wet areas

Yardsticks like scoring systems would assist in the selection of highly maintainable building designs In this regard, quantitative means such as life cycle cost approach can form a strong fundamental for developing a scoring system to measure the level of maintainability Therefore, with the basis of life cycle cost, the research has focused to establish:

o a benchmark by which the maintainability buildings can be

measured and compared

o an indicator to clients the extent of maintenance required through

the design details before they are actually built

o a device by which the maintenance costs over the entire building

lifespan can be predicted

2.10 SELECTION OF A SUITABLE MODELLING TECHNIQUE

In attempts to deal with research and design issues associated with buildings there are often difficulties encountered in handling situations where there are more than six variables involved To adequately model and predict the behaviour of building systems demands the consideration of nonlinear multivariate inter-relationships, often in a ‘noisy’

environment To put neural network in perspective for use in this research, it is worth briefly reviewing the techniques available for the investigation of complex multivariate issues that often arise in the design of buildings The number of variables that can be handled in a scientific study can be significantly increased by the well-established engineering technique of dimensional analysis

Analytical techniques have been successful and are valuable in understanding principles where less than optimal designs were acceptable with the advent of digital computers, numerical methods became much more attractive than analytical solutions, as they could handle more complex and realistic situations However, numerical methods have their limitations as well, namely the incapability to account for practical limitations introduced

by factors such as workmanship Moreover they tend to perform well at analyzing a situation but not so well as a designer’s tool for prompt selection of options Furthermore, the number of variables that can be considered is still limited and numerical solutions cannot be obtained directly for occupant’s perceptions or preferences