Risk management is vital when tendering for construction contracts. Risk is described by Atkinson (2001) as the “ probability of an occurrence of a hazard and the magnitude of the consequences”. Consequently risk can be considered as the likelihood of an experience occurring and the resultant effect of that experience if it takes place.
As defined by RICS (2009) risk management is a means of processes where risks are identified, analysed and managed. It is a constant cycle that begins at the pre-tender stage; this means that risk can be priced into the bid and continues after post contract stage. During the different phases of a project, new risks will emerge throughout the contract. Identifying in advance allows quicker mitigation; to reduce impact risk has on the project.
This study has examined the inaccuracy of pre-tender risk management by using both pre-tender and post contract risk registers. As well as risks there were opportunities and these were also recognised in the processes of risk management.
This subject area was chosen to be researched after discussion with line managers and work colleagues, during my 2010 placement year with Sir Robert McAlpine (SRM). This topic was considered as an ideal subject area for a research project because SRM were improving their risk management procedures.
Therefore the plan of the study was to feedback findings, to further enhance the Tender risk management process. The line manager gave me risk registers for two different project types, and explained how SRM assess risk at all stages of a job. The emphasis being on highway construction, as this information was available from my placement and SRM.
1. 2 Rationale
The subject of what to research developed from consultations with colleagues and the line manager at SRM. The line manager was involved in risk management and was working to improve SRM’s pre-tender risk process to make it replicate the post contract phase. This therefore illustrated a requirement in the business and indeed construction for research in to this topic area. The main reason for eventually electing the subject area of risk was the open access of information provided by SRM and the contact with knowledgeable personnel in my placement. This gave a better understanding of the processes currently used, which allowed me to gain a greater grasp of the subject area. Also access was gained during placement to potential data in the form of contract and tender risk registers from various Highway works undertaken by SRM.
The aim, objectives and hypothesis came about from carrying out the literary research, it wasn’t until that point that the information was collected, and a plan formulated of how to use it. The research aimed to understand how the pre-tender risk process may be inaccurate. After that considering how it could be improved and whether it was possible to do so. Subsequently, recording any conclusions of pre-tender process in order to develop and better it.
1. 3 Aim, Objectives and Hypothesis
1. 3. 1 Aim
To see whether the post contract risk process identifies significant additional risks that the pre-tender risk process failed to identify and thus determine how inaccurate the pre-tender risk process is.
1. 3. 2 Objectives
Identify processes currently used to manage risks at post contract and pre tender stage in recognising and measuring risks.
Analyse whether the key additional post contract risks identified were included at pre-tender stage.
Compare SRM’s risk management processes with other contractors and analyse to find the best solution.
If possible, note any resulting conclusions and input results into the pre-tender risk process to produce a new accurate tender risk management.
1. 3. 3 Hypothesis
The post contract risk process in Highway construction successfully identifies additional risks and as a result pre-tender appraisal is inaccurate.
1. 4 Overview of work done / Methodology
1. 4. 1 Literature Review
In order to understand the processes used to manage risk in the industry a literary review was embarked on. This meant the research would gain an appreciation of how risk is perceived by the industry, as prior to this my personal experience had only come from SRM’s perspective and the processes they use. In order to gain real understanding and to funnel clearly on what my hypothesis stated, it was decided to split the literature review into two sections. One section, comprising of processes used by industry, taken from an assortment of books. The second section being that of SRM’s highways process and accompanying documents. This makes it possible for an assessment to be conducted between SRM and the other contractors, so as to gain a better view of how Risk management at different stages of a contract work in Highway works.
1. 4. 2 SRM Projects
The 3 SRM projects looked at were M1 J25-28, A19 upgrade (both Highways Agency) and M74 Glasgow council, they were all Highway maintenance contracts although they differ slightly. A19 was Term Maintenance Contract (TMC) whereas M1 J25-28 and M74 upgrade were (MAC) contracts.
When examining the A19 upgrade the contract risk register was acquired but the tender risk register for this contract was not. However the contract risk register for this project was compared with a tender risk register from another contract. Not ideal but the only solution at the time. The majority of the risks are common and on most projects registers, so comparable/related risks were drawn from the tender register, and included in the A19 contract risk register.
An evaluation was done involving the contract risk costs, and tender risks. This was done by incorporating the maximum, minimum, most likely costs and the probabilities from contract and tender appraisals. A total assessment for each risk occurred via averaging the maximum, minimum and most likely costs then multiplied by probability of risk occurrence. All risks types were identified, in order to distinguish trends in risks identified at pre-tender stage and found at post-contract stage or were not found at all.
The M74 contract risk register was distinctive to M1 J25-28 and A19 upgrade registers as it didn’t contain costs. Every risk was divided by its probability and impact, whether that was high, medium or low assessment. Having a pre-tender risk register for this project, meant risks were matched up with the contract risks. Assessing if each risk was covered was done by assigning costs and probabilities to every risk. To find the overall assessment, the probability and cost were multiplied. These costs were then evaluated against the overall pre-tender costs, as with the A19 upgrade.
1. 4. 3 Interviews
Interviews via telephone were conducted with a few companies to really see what processes are used at post contract and pre-tender. Also to find out how precise these methods are. Companies interviewed were:
- Morgan EST.
- Vinci
- Balfour Beatty
- May Guerney
1. 5 Overview of main conclusions
The data and analysis carried out imply there is data to sustain in part the hypothesis. This is due to a large number of risks being identified in contract stage that were not previously seen at pre-tender. Even though assessments for any one risk were fairly inaccurate, the overall assessments for pre-tender and post-contract stage were close. Showing that the pre-tender risk process is inaccurate and needed improving if assessments and risk management is to become more reliable.
1. 6 Guide to the report
Chapter by Chapter overview of report:
1 – Introduction – An opening on risk management in relation to the project and validates the basis for choosing the subject area. It also states the aims, objectives and hypothesis which the project is established upon, as well as outlining the work done and an overview of the Conclusions.
2 – Methodology – Explains the methods used to investigate my hypothesis, from Literature review, approaches used data collection, analysis and interpretation of results. This describes the reasons for using the methods chosen and any research boundaries/ limitations.
3 – Literature Review – This chapter contains my understanding and background reading for the subject area. This was done by, studying the risk process used in management. Allowing a greater comprehension of risk in Highways work, and how it’s viewed and used.
4 – Results and Analysis – Confirms the results of the research, and the subsequent analysis for the SRM projects and other contractor’s interviews. It outlines the assessment made concerning the pre-tender risk register and the post contract risk registers. From this it then cross-examines the data so as to be able to test the hypothesis.
6 – Conclusions and Recommendations – This analyses all the results in relation to the hypothesis and whether they support it or not. It details any limitations that affected the project, while also imparting proposals for both industry and any future dissertations.
2 Methodology
2. 1 Introduction
The methodology was vital to the accomplishments of research and was dealt with care in order for the most appropriate research methods to be chosen (Fellows and Liu, 2008). The methodology outcome depended on the subject area, research aims and amount of literature review obtained. For data collection and analysis, the methods employed ought to be realising the aims and objectives so as to continually test the hypothesis and validate the research.
The information that was obtainable and available played a huge role, as work on the dissertation could not be done if the information wasn’t relevant. The information therefore can be decided by the hypothesis, as if the information is not on hand then one cant trial the hypothesis. Due to these factors, risk management was identified as an appropriate topic from the beginning, but it wasn’t until the research was started that defining the hypothesis was possible, yet having an impression of the aims and objectives that were to be accomplished. From the off, the aim was to gauge how precise pre-tender risk management was, even if unsure of the data and information existing prior to consulting SRM team.
Ultimately this section highlights the research methods utilised in the research, and the close association that has been made when doing so between the methods and Aim, hypothesis and objectives.
2. 2 Aim
Employed to help concentrate the methodology in choosing the right methods, also to clarify to the person who reads the dissertation what precisely is being investigated. So this shall be done by recalling the Aim set out:
To see whether the post contract risk process identifies key additional risks that the pre-tender risk process failed to identify and thus determine how inaccurate pre-tender risk process is.
2. 3 Quantitative and Qualitative methods
In Data collection there are two key styles; quantitative and qualitative forms of research. Quantitative being the collection of data measured with figures and analysed with statistical trials in order to to test the hypothesis (Creswell, 1994). While Qualitative research is quite different, it is a method that uses meanings, experiences and descriptions (Naoum, 2007). Quantitative research can be easier to examine as it creates measurable/quantified outcomes that can create analytical arithmetical results. While, qualitative information from research has a tendency to be complicated as it often requires researcher input and manipulation to ensure it’s appropriate for investigative procedures (Fellows and Liu, 2008). The research methods depend on the information obtainable and the aim of the research. In carrying out good piece of research, it’s generally required and essential to use both quantitative and qualitative. A combination of methods was used, to enable the correct and more reliable conclusions; more is detailed further in this section.
2. 4 Literature Review
A literature review was undertaken; to provide the core subject knowledge of risk management in construction. The idea was to comprehend how risk is analysed in the construction sector, and the procedures used in controlling it. Next research was done to discover how risk is managed, by using literature in form of books and journals. Then see SRM’s risk management process from their risk management documentation.
Dividing the literature review in two sections meant that one of the objectives could be fulfilled by comparing SRM’s risk management approach with other construction companies.
Overall the literature review gave a greater understanding of the chosen topic of risk, and illustrated the problems and successes in risk management. While also showing the diverse and numerous ways in managing risk, and how the approach identified and selected can depend on many factors such as project size, contract used and size of companies involved in the managing of the project.
The list literature sources below were used in delivering the dissertation:
Textbooks -were very useful in gaining the relevant knowledge of risk management, and procedures used. Books were found by searching Loughborough University’s library database. Chapters needed or thought appropriate were studied, and compared against searches that had been made on the internet. A check system, which assessed the validity of both sources of information against each other. With much of the research it became apparent that information in books available weren’t current especially in the older series of books. Despite this it wasn’t a predicament for Risk management as texts of up to ten years old were and are significant and applicable now, with some techniques having improved. The books allowed for great comparison for up to date information on the internet.
Journals – were again located on the Loughborough University library database, which with the relevant buzz words located material of use. Finding journals proved difficult in comparison to textbook numbers, yet the sources were helpful.
Internet – supplied a platform for research of literature. Being easy to use and handy, collating information could be done at speed and with relative ease. It allowed greater understanding of what type of book would be needed from the library. Information from the internet was important but it had its limits and it was vital to know that it can be inaccurate and cause misguidance, research can be more guaranteed with text and journals. This way of thinking about limitations in using the internet was in mind when searching websites on risk in the construction. Finding various helpful websites any information was compared other websites in order to increase reliability, but most importantly against book and journals. As with all research appropriate sources of reliability were identified and used like the RICS website. Two editorials from RICS site provided constructive, and were used in the literature review. In finding books the internet was most useful as many articles on websites highlighted book of particular relevance and use.
SRM’s Risk Management Procedures document – SRM’s risk management procedure document was used as section two of the literature review. Reading through the document and important information was used in the literature review, particularly, on the processes used by SRM in risk management. A comparison was made between SRM’s procedures and with those found in the first half of the literature review.
2. 5 Data Collection
When actually collecting the data for the research the collection was again split into two parts. The initial data collection was from SRM projects, other data collection was associated to other contractors. The intention was to compare the risk procedures, and attempt at analysing which was the superior one. The next section shows how all the data collection was collected and prepared.
2. 5. 1 SRM Projects
The data for SRM projects was collected during a placement year, when working in the relevant packages and job roles. Prior to any data collection, discussions with line managers and seniors took place in order to help formulate and aid the research as what would be required in terms of data. At this stage no hypothesis had been decided, the idea was to assess the accuracy of pre-tender risk assessments. It was therefore recommended that the applicable information in the form pre-tender and post contract risk registers could be provided. The contracts that would be made available were the M1 J25-28 scheme, A19 upgrade and M74 Glasgow project. Three similar Motorway/road maintenance projects that differ from each other contractually.
A19 upgrade is a TMC to maintain, operate a network of strategic roads in the North East. M1 J25-28 and M74 are both MAC contracts for similar maintenance one in the midlands and the latter in Glasgow. A19 upgrade and M1 J25-28 are HA run, while M74 is for the Glasgow city council. The HA run their contracts by splitting their contracts into 13 sections in the UK, and in these projects cases offer for the extension and maintenance of the roads within these sections.
The reason that A19 upgrade was a TMC but M1 J25-28 and M74 MAC contracts is previous to the MAC HA projects were also run with the TMC contract, but now all is done under MAC style. The contracts run for 5 years with optional extensions. Different sections of maintenance start and end at different areas on the motorway in question over a 5 year cycle, meaning that when MAC contracts were first used, some sections weren’t using them and using TMC etc…
A19 upgrade was a TMC’s completed in 2009, whereas M1 J25-28 scheme was a mew MAC contracts just seeing completion in late 2010. The difference in TMC’s and MAC contracts is TMC’s two separate companies, one as managing agent the other as main contractor. MAC contracts are one company, who runs as both the contractor and managing agent.
Using the three contracts as the basis for the research the risk manager on the placement at SRM started off by providing a tender risk register for the M1 J25-28 and A19 contract, projects he had involvement on. Trying to find tender risk registers for M74 was problematic but couldn’t find any. In order to have a full complete analysis, it was decided that for M74 to use another MAC contract tender risk register. This was exactly the same as M74 except the location geographically. For sure this would create limitations but it was decided, it would possible to use the risk register as SRM tend to re-use the risk register from preceding bids for specific contracts. So similarities would be high and that as long as it was noted in the dissertation as to its use and reasons why. Due to the generic nature of the risks they were deemed suitable as with what SRM do in practice.
While on placement working on the M1 J25-28 contract meant developing contacts with people in the relevant field of risk management. Therefore requesting the use of the risk register on M1 job was easier than the other projects. The A19 contract risk register was also obtained contracts at work so. Finally also getting the M74 contract risk register, because although never having worked on this contract the line manager on my placement gave contact details of relevant personnel to enquire with and the registers were duly emailed.
2. 5. 2 Other Contractors
Obtaining data via contractors in the sector of road maintenance was much more difficult than from SRM sources, as they were very protective of information they gave out. All the projects obtained so far from SRM contracts were Highways related, so the focus was on trying to collect data from Highway contractors. By doing this it would allow for a honest comparison between SRM and other contractors. Processes to manage risk are expected to be similar involving different construction projects; the data was likely to be different.
The plan was to send questionnaires to the chosen contractors, but after unsuccessful returns it became apparent that another route would be needed, so interviews via telephone were deemed suitable. It seemed with written questionnaires, companies were more likely to ignore them, whereas on the phone they would respond to the questions asked immediately, with no real confusion of what the question was asking as myself in person could explain.
The Highway contractors contacted were:
- Morgan est.
- Vinci
- Balfour Beatty
- May Guerney
A semi-structured set of questions was planned in order to gain the information required but also allowed telephone participant to elaborate and discuss the subject in a friendly professional manner. Contractors in Highways works were contacted, the dissertation of Risk Management was explained with the aims and objectives that needed to be achieved and the following questions were asked:
- Do you have a pre-tender risk management method?
- How do you detect risks in the pre-tender phase?
- How do you calculate risks to arrive at a total risk pot?
- Do you undertake risk management post-contract phase?
- Could I obtain a pre-tender risk analysis and post contract risk analysis for my research use (for the Highways project you are on)?
To each conversation it was explained that my aim was; compare the risks types at tender stage with contract stage as well as assessments made. The questions were supposed to identify the processes companies used in managing risks. Overall the different companies were helpful in answering these questions. The Fifth and final question was done to get registers like that gained from SRM. It was the only difficult part as many refused to hand registers over from live projects due to the sensitivity of their data and company policies. However some registers were received but didn’t really contain the crucial information required just a formatted company risk register. Throughout the interviews, notes were made on the first four questions and the results put into a table at the time of the interview showing the company and their response to each question. Thus, making analysis easier when looking at the responses later in the dissertation.
2. 6 Data Analysis
Prior to data analysis, the research had assessed how SRM price risks because of the literature review and looking at the risk registers. It was crucial to identify this before undertaking the data analysis as it established what type of analysis would be carried out, and therefore detailed below.
Upon formulating the risks that have been identified each risk will be categorised with minimum, maximum, most likely value and probability. Done for all of the risks, the information is put into a piece of simulation software called @ RISK which does hundreds to thousands of simulations, and creates a graph with a bell-shaped curve.
The 75th percentile is put in the bid as the risk potential. SRM do this for all of their projects. For this dissertation we were concerned in the risks were identified and the individual costing of each risk as shown in paragraph above with min, max etc. The focal point being what was keyed in to @ RISK as opposed to what it produces.
For the Data analysis the SRM contracts have been split into their individual projects and the Other Highway contractors. The reason for splitting the SRM’s projects is that the analysis varied in parts by way it had been collected as mentioned earlier (some with full risk registers some in part).
2. 6. 1 M1 J25-28s scheme and M74 Glasgow
Analysis for M1 and M74 projects were pretty much the same. The M1 contract and tender risk register could be compared with the contract and tender risk register for the M74 contract, due to them both being Mac’s.
To evaluate the accuracy of the risk assessments, comparable risks to the contract register, were removed from the tender register, put against the matching risk in the contract register. When comparing values in matching risks, it was vital to have a total cost for risks in both the contract and pre-tender register.
The best and most impartial way to do this for the pre-tender risk register was to take the average from the maximum, minimum and most likely figures and multiply the probability. The maximum, minimum and most likely values are the range of potential expenses that could be incurred by the risk, and any total cost was decided to be an average of these, as all projects had them. As its unknown as to it’s actually occurrence the average should be multiplied by the probability, which gives a total potential cost to the risk.
While the contract risk registers should be considered by severity and possibility by scale of 1-5, and then an assessment of the risk is undertaken, figuring out the minimum, maximum and most likely values of each risk.
Mitigation measures are identified for each risk, and then assessment is done again as before the mitigation. The M74 contract risk register did not have minimum cost of the risks, so all risks were assumed to be zero. Without any minimum values any overall assessments of the risk would not have been made. The reason for not using the likely value instead was so the data would use a range of values. It was decided greater accuracy would be found in setting the minimum as zero and range of values than the most likely. By setting all its risks as zero means they are an unimportant minimum value or a risk/event that doesn’t happen. The M1 contract register had all the necessary values so no intervention was needed.
Both contract risk registers failed to show any probability, just a likelihood scale from 1-5. Therefore it was assumed, giving the scale a percentage as would normally be done:
1 – 10%
2 – 30%
3 – 50%
4 – 70%
5 – 90%
These percentages were used because they provided a suitable range, as risks with low likelihoods (1) are unlikely to occur but not impossible so 10% seemed a reasonable percentage. Similarly 90% seemed a reasonable percentage for high likelihoods (5) as they are likely to occur but not certain. The other values were then evenly distributed between 10% and 90%. Having made these assumptions the overall assessment for contract risks was made in the same way as the tender risks, finding the average of the maximum, minimum and most likely and multiplying this by the probability. Where the same tender risk was identified as being applicable to more than one contract risk the overall assessment was divided by the number of contract risks it was applicable to. This was because when these values were totalled there would be double counting of these tender risks if this was not done.
To show this information a table was created showing; a list of contract risks, the corresponding tender risks, raw data inputted into the risk registers, and the overall assessments. I then split the contract risks into one of the following categories:
Not identified but covered (No cost).
Not identified and not covered – should be identified.
Not identified – cannot be identified.
Identified and covered.
Identified but not covered.
Each risk was grouped by colour to state which category from above it fell in.
2. 6. 2 A19 upgrade
The analysis for the A19 contract was started off in the same way as had been done for the M1 J25-28 and M74 contracts, going through the contract risk register and identifying any similar risks from the tender risk register.
The assessments for the overall cost for the tender risks were made in exactly the same way, by taking the average of the maximum, minimum and most likely values and multiplying this by the probability. This was because the tender risk registers were in exactly the same format. However the differences came when the assessments of the overall cost for the contract risks were made. This was because for this project, the contract risk register assessed the risks in a different way to the Area 6 and Area 13 contract risk registers. This difference was that there were no costs in the A19 risk register.
Instead the risks identified at contract stage were assessed in terms of likelihood and impact on a scale of high, medium and low. This meant that assessing the overall cost for the contract risks were harder because there was no costs given. Therefore the only way to assess the overall cost of the risk was to give the risk a cost and a probability based on whether it was high, medium or low and then multiply these two figures together. This means that deciding what costs and probabilities to assign to each level of risk was important, as the overall assessment was dependent upon these assumptions. In terms of what costs to give for each level of impact, the risk matrix that SRM use for prioritising risks was referred to. They assess the impact and probability of each risk using a 1-5 scale and they give the parameters for impact as being:
1 – under £1, 000
2 – £1, 000 – £10, 000
3 – £10, 000 – £100, 000
4 – £100, 000 – £1, 000, 000
5 – Over £1, 000, 000
As this was a 1-5 scale and the risks and the contract risks were only divided into high medium and low, the figures for 1 and 5 was as too extreme both ways. For low risks I decided to take the high point of a risk impact of 2 (£10, 000) and for high risks I took the low point of a risk impact of 4 (£100, 000). For medium risks I took the midpoint between these two values (£55, 000). Using these figures seemed reasonable based upon this scale, as it created enough of a range without a too big range.
For the probabilities, the likelihood scale they use was based on descriptions rather than probabilities. From knowledge in research the probabilities for low were set at 10%, medium risks 50% and high risks 90%. These percentages were used because they provided a suitable range, as risks with low likelihood’s are unlikely to occur but not impossible, so 10% seemed a reasonable percentage. Similarly 90% seemed a reasonable percentage for high risks, as they are likely to occur but not certain. For medium risks the midpoint between these two percentages (50%) was used, because they are possible to occur. The cost was then multiplied by the probability to give an overall value for each risk. Following this, the rest of my analysis was exactly the same as the M1 J25-28 and M74 contracts
2. 6. 3 Further Analysis & Explanation
This section so far details how the risk registers were compared, which was the first step in terms of analysing the data, and these tables are included in the appendices at the end of this dissertation.
However on there own, these tables did not give sufficient information to be able to test my hypothesis. Firstly, to interrogate how accurate the identification of the risks was, the risks that were identified, and were not identified and not commercially covered, were filtered out, and lists were made of these risks. As this was done, each risk was put into a category to see if there were particular categories that are, and are not identified at tender stage. To analyse this, tables were created for risks identified and not identified, detailing the categories of risks, and the number of risks in each category. From these tables, two pie charts were drawn to show this information graphically.
To analyse the assessment of the risks, a summary table was firstly drawn to show the number of contract risks in the following categories, and the total assessments relating to these risks:
Not identified but covered (No cost).
Not identified and not covered – should be identified.
Not identified – cannot be identified.
Identified and covered.
