Research Report

5.1. Collection and Analysis of Information using the Grounded Theory Method

As described above, we used three methods to collect data:

• A workshop with the Steering Group,

• Structured Interviews with the members of the steering group and a number of other companies in the sector;

• Collection of quantitative data relating to individual error occurrences. This included:

- A descriptions of each of the errors,

- A description of the causes of each of the error occurrences,

- An assessment of the financial cost of each of the error occurrences.

This information was collected from the participants in the structured interviews. The Steering Group workshop and the Structured Interviews were very effective and provided useful data. The collection of quantitative data was less effective. Few of the organisations that we interviewed had detailed quantitative data relating to errors. Where information was available the financial details generally related solely to the direct cost of error to the organisation being interviewed, rarely was information available on the cost (direct or indirect) to other parties. The quantitative data that was available all related solely to the direct cost of recorded error. Although the information relating to individual error occurrences was limited, we were able to use the data that was provided about turnover and the reported total direct cost of error to individual organisations to review estimates of the total direct cost of recorded error. 

Tier 1 Contractors have an overview of the construction process and provided useful information about the incidence and impact of error across the different trades. The initial cost of errors is incurred by the Tier 2 Contractors and this cost is often invisible both to Tier 1 Contractors and to end Clients.

5.1.1. Areas of work in which errors occur with greatest frequency

The limited availability of quantitative data has made it impossible for us to identify the errors that have the largest financial impact in the UK construction industry.

We have been able to identify the areas of work in which errors occur with greatest frequency, as reported by our study group. The ranking in the table below is by the reported frequency of occurrence of error. The ranking does not necessarily reflect the economic significance, as some errors whilst frequent may have low cost impact.

1. In Situ Concrete
2. Drainage
3. Piling
4. Mechanical Systems
5. Finishes
6. Damage to completed works
7. Facades / Cladding
8. Electrical Systems
9. Roads and Pavements
10. Roofing
11. Controls /
12. Timber
13. Setting out
14. Steelwork coatings
15. Temporary works
16. External works
17. Structural steel
18. Lifts
19. Cold bridging
20. Damage to live services
21. Underground waterproofing
22. Utility connections
23. Tunnels
24. Masonry
25. Earthworks
26. Basement waterproofing

Timber errors were reported have a relatively higher frequency but lower financial impact.
Errors in setting out were reported to be low frequency but high cost.
Some respondents reported a very high rate of error in steel coatings with a high knock on effect on following trades. Although reported to be areas of work in which there is a lower frequency of occurrence of errors “Underground Waterproofing” and “Basement Waterproofing” were reported to be areas of work in which the financial impact of error is significant.

5.1.2. The root causes of error

The list below sets out the root causes of error that were reported within our study group. The ranking here is by frequency of reporting. The ranking does not necessarily reflect the economic significance as some of the root causes, whilst frequent, may have low cost impact. It was reported that although the errors and defects vary between trades, the causes of the defects are often the same.

1. Management and Planning
2. Communications
3. Workmanship
4. Design
5. Materials
6. Interface Design and Management
7. Commercial Pressures
8. Late changes
9. Information overload
10. Inadequate training
11. Setting out
12. Fabrication
13. Lack of focus on quality
14. Designers having a poor understanding of details
15. Design fees
16. Late surveys and investigations
17. Programme
18. Time
19. Core skills
20. Outside of core geographical range of work
21. Motivation

Management and Planning

Lack of planning at all levels (including operatives) is a key cause of error. There was a view that planning skills are not as good as necessary particularly when planning involves co-ordination across trades. There are often valid reasons why work does not proceed in accordance with the plan. Once the works start to deviate from the plan matters quickly go from bad to worse.

There is often insufficient supervision of site works. Supervisors are often insufficiently briefed or trained on how to supervise the trades and techniques on progress on that day: The same person may be involved in supervising a wide range of different trades.

Scarce resource for supervision of site works is often poorly allocated to areas where the most work is visible, rather than to the areas where there is the greatest risk of error. Works that are difficult to access and inspect often suffer from a lack of site supervision and inspection.

There is often insufficient checking as the work progresses; the consequence is that the financial cost of errors is increased. Poorly aligned package scope and trade contractor skills can result in error. This is particularly the case when specialist trade contractors are required to manage further subcontractors. There is a general determination to get it done so that subsequent works can proceed. Sometimes it appears easier just to do something than it is to take the time to work out what is the right thing to do. All of the study group have talked about problems arising because of a failure to make time to plan or to find out how the work should be done.

As an industry we have a real “can do” approach. We are instinctively keen to get on with each stage of a project and often proceed before the work has been fully planned. This “can do” approach is present across the industry, ranging from a tradesman who starts work before he or she has fully thought through what they should be doing to construction work started before the design is complete (thereby risking design changes with all the costs and disruption that can incur).

We never have time to get it right. We always have time to put it right. It is sometimes difficult for staff to ask for help. There is a lack of awareness that one person cannot be an expert about everything.

Communication

Communication between all parties was reported to be a key cause of error. This included all forms of communications: written, drawn and verbal.
Communication ‘up the chain’ can be a particular problem:

Site operatives sometimes lack the confidence or motivation to ask questions or make
comment. Sometimes site operatives are insufficiently heard.

Site culture or relationships sometimes make it difficult for suppliers to discuss in good time
the possibility of a missed delivery date. Many examples were given of people not being able to express uncertainty or ignorance to their superiors or workmates. Task Briefings sometimes concentrate on the safe way of doing the work and fail to address the technical requirements. Documents, particularly specifications, are often too long and contain much irrelevant information. Irrelevant information is often left in because the person preparing the document lacked the experience to know what is unnecessary – or else information is left in “just in case” which makes it difficult to identify the information that is relevant.

Some respondents were sceptical that BIM provides real benefits.

There is a lot of emphasis in method statements as to how to do the work safely but there is rather less so on how to do the work without error.

Workmanship

It was widely reported that errors in workmanship are only rarely a result of a lack of skill on the part of the site operative. Workmanship and operative technical ability are not generally seen as one of the principal causes of error, although there are isolated cases of poor workmanship or technical ability that do result in significant error. For example, it is not generally felt that errors in brickwork arise because bricklayers are not capable of mixing mortar and laying bricks, rather that for one reason or another they do not do what they are capable of doing consistently enough.

From the discussions it was understood that the errors assigned to ‘workmanship’ relate generally to failures in planning the task rather than to technical competence. A view that has come up repeatedly is that people lack sufficient contextual knowledge to be able to complete their work effectively. In particular a lack of proper understanding of the roles undertaken and challenges faced by others involved in the construction process often results in inappropriate decisions or actions being taken. The study group have an almost endless supply of examples where errors have arisen as a result of this mutual ignorance. A few examples are given below as illustrations of situations that appear to be commonplace within the industry:

In ignorance a designer unnecessarily specifies fixings from a manufacturer’s catalogue that are not stock items, resulting in procurement delays and consequent changes to planned construction sequence.

The procurement team substitute the specified 200m of 150mm diameter clay pipe with 200mm of 150mm diameter plastic pipe (to reduce cost) without understanding that the ground on the site in question has relatively high levels of hydrocarbon contamination and that the plastic is notsuitable, with the outcome that the Contractor has to dig up and replace the entire drain run.

The bricklayer uses 10.5N 100mm blocks to build the non-load bearing partitions which only require 3.6N blocks not realising that they are a relatively long lead item specifically intended for the load bearing walls nearby. The outcome in this case was delay and additional costs incurred to replace the 10.5N blocks.

Interestingly in none of these examples did the error result in a defect, although all were relatively costly for the projects concerned.

Design

Design Changes (late or otherwise) were identified as a key cause of error. Designs, and detailing in particular, are sometimes unsuitable.

Materials

It was reported that errors arising from problems with materials were often a result of late changes. It was reported that late changes in materials are often made by people who do not understand the effect of the change being made – often the person making the change is unaware that they have ‘made a change’. For example, it was reported that buyers sometimes order alternatives without being aware that there are real technical differences between the more expensive product specified and the cheaper alternative. Materials are sometimes changed because the item that was specified cannot be supplied in the time that is available. 

Interface Design and Management

It was reported that, as well as the design and coordination of interfaces, tier 2 Contractors are often put under pressure by late handover from previous trades which contributes to errors. It was reported that when a tier 2 contractor performs well, they may be inundated with work without adequate consideration of capacity and that this does result in errors.

Several of the study group have identified poor management and planning associated with interfaces between different systems and trades as being a particular problem.

Setting out

Although this was reported to be a cause of error, we consider incorrect setting out to be an error in itself with the costs arising being an indirect cost of the setting out error.

Fabrication

Although this was reported to be a cause of error, we consider incorrect fabrication to be an error in itself with the costs arising being an indirect cost of the setting out error.

5.1.3. Capturing the cost of error

There is significant variation in whether and how companies collect and report the costs of errors and defects. Most companies recognise that there is a cost issue with errors, but it appears that none is clear as to the full extent of those costs.

Few of the organisations that we interviewed had detailed quantitative data relating to errors. Where information was available the financial details generally related solely to the direct cost of error to the organisation being interviewed, rarely was information available on the cost (direct or indirect) to other parties. All of the systems that we saw only capture the direct cost of error, and do that only partially.

None of the organisations that we interviewed was able to provide data relating to indirect costs or unrecorded process waste, although some participants were willing to express a view.

The direct cost of errors which result in defects (a defect is any failure to meet the project requirements at a handover). Where Contractors do record the results of error the records generally relate to failures to meet the project requirements at handovers – “defects”.

Some of the organisations that we interviewed have detailed records of pre-completion defects while others have details of post completion defects. None of the organisations that we interviewed had information available on both. It was reported in some organisations that the board was principally interested in the cost of latent defects since these were perceived to affect ‘the bottom line’.

Several organisations had systems in place which did not focus on cost of error, but instead recorded the number of defects remaining on completion or, in different organisations, the number of defects reported and resolved during the construction process. None recorded everything.

There was generally more formalisation of the error recording systems in organisations with a larger proportion of large publicly funded projects. Most contractors acknowledge that the recording systems which they use do not allow them to capture the cost of error in some areas. For example, some Tier 1 Contractors reported that the costs of M&E defects are not disclosed by Tier 2 Contractors.

There was insufficient standardisation in the recording of error in the study group for us to use the detailed quantitative data which we did receive from a number of the study group to make a meaningful assessment of the cost of individual error types.

The bulk of the direct cost of error relates to materials, plant and labour costs. These costs are carried by the Tier 2 Contractors. We were not able to obtain good data on the direct cost of error to Tier 2 Contractors.

The direct cost of error to Tier 1 Contractors is largely associated with the additional management associated with rectifying errors. A review of the data from the Tier 1 Contractors suggests that the direct cost of managing errors which result in defects is in the range of 0.5% – 1.0% of project costs. Not surprisingly the total direct cost of errors which result in defects to Tier 1 Contractors with significant direct labour costs is reported to be substantially above 1.0%.

Tier 1 Contactor costs for managing construction projects are generally in the order of 10% – 20% of the total cost of construction (depending on sector and project type). Therefore, the reported direct cost of errors which result in defects to Tier 1 Contractors falls in the range 2.5% (0.5% / 20%) to 10% (1% / 10%) and is probably typically around 5% of turnover.

The direct cost of errors which are resolved before a handover – “unrecorded process waste”

Very few Contractors record errors and the associated rework and extra work costs that are incurred before handovers. The direct costs of errors which are resolved before a handover are very rarely recorded. Most of the direct costs of errors which are resolved before a handover are incurred by Tier 2 Contractors and below.

It is likely that only a relatively small proportion of the total direct cost of error arises from errors which result in defects. The majority of the direct cost of error is understood to arise from errors which are resolved before a handover and which are not recorded as defects.

Therefore, the majority of the direct cost of error is, in effect, experienced as “process waste” at Tier 2 and below. We were not able to obtain sufficient quantitative data relating to unrecorded process waste to report a cross industry figure.

When working with the NEC form of contract Tier 1 Contractors are sometimes rewarded if “NonConformances” amount to less than 2% of the contract value. In these cases the relevant “NonConformances” are items identified at handover by the Client’s Representative. We have been informed that the results of the majority of errors are usually corrected earlier, after having been identified during inspection by the Tier 2 Contractor or the Tier 1 Contractor. This indicates that the direct cost of error is generally substantially greater than 2% of the project cost.

The indirect cost of errors

Generally, Contractors were not able to provide an estimate of the indirect cost of error. One respondent estimated the indirect cost of error to the organisation to amount to an uplift of around 40% on the direct cost of error. 

The total cost of errors

No Contractors were able to provide an estimate of the total cost of error. All respondents agreed that the cost of error was substantially greater than the recorded cost. Few were able to provide an informed estimate of the total cost of error. Respondents told us that they believe the cost of error to Tier 2 Contractors in some trades to be between 20% and 25% of turnover. The respondents were not able to provide evidence to support these statements. Based on the discussion above it appears highly probable that the total cost of error in construction is considerably greater than 10% of the total cost of construction.

Tier 1 Contractors have very little understanding of the true cost of error which is borne by their supply team. It appears perhaps not surprising that Clients are largely unaware of the amount of error embedded in the construction process or the consequent costs to them of these errors. There is some evidence to suggest that whilst the initial costs associated with errors occur in Tier 2 and below, the root causes may sit outside their control.

5.1.4. Methods used for avoiding error and minimising the consequences of error

Our research has also identified that most Contractors have a quality management system which is, in part,intended to reduce error and the associated financial impact. Several different theoretical frameworks have been used as a basis for these quality management systems and there does not appear to be an agreed industry standard.

Although different Contractors use different systems, the analysis demonstrates that adoption of systems designed to reduce error can deliver substantial financial savings.

In general Contractors that focus on large public sector projects have a more developed system for recording and assessing the results of error. Contractors working on smaller private sector projects are, in general, likely to use a less developed system or to have no formal system in place.

A number of companies have some excellent methods for eliminating error. These included quality circles, encouraging feedback from the workforce, benchmarks, use of mock ups and prototypes, encouraging people to stand back and think, encouraging people to ask when they are not sure, and quality risk assessments. The best examples of Good Practice come from those companies that have effective senior leadership interest in quality and that build an effective culture. 

Among the people that we interviewed there is a strong school of thought that if you get the culture right everything else falls into place, including a reduction in the occurrence of error. It was emphasised that leadership is critical to definition of the culture of an organisation.

A demonstrated interest in quality by a Company’s Board has a positive effect on reducing defects, markedlymore so than a market approach which assumes that defects will be eliminated by focussing on costs.

Sites which adopt practices which result in high quality with few defects are generally safe sites almost certainly because there is good management. It was reported that fewer errors occur on projects where the Client has put in systems to monitor quality. It was also reported that there are fewer errors on projects where Clients demonstrate an interest in quality. Independent checks completed by an external audit organisation were reported to be an effective means of reducing error.

5.2. Analysis of the Grounded Theory results using the Delphi Method

We used the Delphi Method to rank the results of the Grounded Theory Method analysis and to assess the relative financial impact of:
– The direct costs of error,
– The indirect costs of error,
– Latent defects and
– Unrecorded process waste.

The Delphi Method was also used to identify the most effective methods of reducing the financial impact of error in the construction industry.

The results of the analysis using the Delphi Method are summarised below. The changes in the results between the first and second round of the Delphi Method analysis were small.

5.2.1. Ranking of the areas of work according to the financial impact of error

The analysis using the Grounded Theory Method provided us with a list of the areas of work in which errors occur with greatest frequency, as reported by our study group. The ranking does not necessarily reflect the economic significance, as some errors whilst frequent may have low cost impact. There was some overlap between the categories with resulted from the Grounded Theory Method. 

For the Delphi Method analysis we required a concise list of the areas of work. The table below summaries the consolidation of the categories and the amended category titles which were adopted for the Delphi Method analysis. Results of the analysis using the Grounded Theory Method Categories taken forward for analysis using the Delphi Method to identify the areas of work in which error has the most significant financial impact. Areas of work in which errors occur with greatest frequency, as reported by our study group.

Notes:
A. The purpose of the analysis using the Delphi Method was to identify the areas of work in which error has the most significant financial impact. Although reported to have a high frequency, timber errors were also reported have a low financial impact. To allow the experts to focus on the key issues this category was therefore omitted from the analysis using the Delphi Method.

B. The purpose of the analysis using the Delphi Method was to identify the areas of work in which error has the most significant financial impact. Following the structured interviews it was understood that these areas of work in which errors occur at lower frequency were not the areas of work in which the financial impact of error was most significant. To allow the experts to focus on the key issues these categories were therefore omitted from the analysis using the Delphi Method.

C. Although reported to be areas in which there is a lower frequency of occurrence “Basement Waterproofing” and “Underground Waterproofing” were reported to be areas of work in which the financial impact of error is significant. Therefore these categories were combined and included in the analysis using the Delphi Method. The first question of the Delphi Method analysis required each expert to rank the areas of work according to the financial impact of error in each area. The results of the analysis using the Delphi Method are summarised below. The changes in the results between the first and second round of the Delphi Method analysis were small.

5.2.2. Ranking of the root causes of error according to financial impact

The analysis using the Grounded Theory Method provided us with a list of the root causes of error that were reported within our study group. The ranking is by frequency of reporting and does not necessarily reflect the economic significance as some of the root causes whilst frequent may have low cost impact. There was some overlap between the categories which resulted from the Grounded Theory Method, and the review of the results indicated that some of the categories would be better separated into a number of sub-categories. For the Delphi Method analysis we required a concise list of the root causes of error. The table below summaries the adjusted categories and the amended category titles which were adopted for the Delphi Method analysis.

5.2.3. The relative financial cost of error

In the third question of the Delphi Method analysis the experts were asked to assess the relative financial impact of: the direct costs of error, the indirect costs of error, latent defects and unrecorded process waste. The estimated distribution of the costs was similar for the Civil Engineering and Building sectors. The results are summarised in the table and chart below.

5.3. Online survey results

There were 143 responses to the online survey. However not everyone that started the survey answered all parts of each of the two detailed questions, the number of responses to each part varied from 59 to 73 of the total 143 participants. The majority of respondents were trade contractors.

5.3.1. Ranking of the areas of work according to the financial impact of error

The online survey respondents were asked the same question that the experts were asked during the Delphi Method Analysis. The respondents were asked to rank the areas of work according to the financial impact of error in each area. The results of the online survey are summarised below.

5.3.2. Ranking of the root causes of error according to financial impact

The online survey respondents were asked the same question that the experts were asked during the Delphi Method Analysis. The respondents were asked to assign an importance of each of the root causes on a scale of 1 to 10. The results of the on line survey are summarised below.

5.4. Comparison of the Delphi Method analysis results with the online survey results

5.4.1. Ranking of the areas of work according to the financial impact of error

The charts compare the rankings assigned in the Delphi Method Analysis with the rankings assigned in the online survey

5.4.2. Ranking of the root causes of error according to financial impact

The charts compare the rankings assigned in the Delphi Method analysis with the rankings assigned in the online survey.

5.5. Literature Review

The literature review is presented in Appendix B. The key findings are summarised below.

5.5.1. Errors

While many organisations maintain rework tracking systems, there is significant variation in the systems used.

There are many ways of describing construction error and the consequences of error. One system used by Reason (1995) usefully categorised errors as:

• Failures of intention where the design or implementation plan is inadequate.

• Failures of execution where actions do not go as planned.

• Deliberate violations where the works deliberately deviate from the plan.

The desk research demonstrates that when we look for the causes of an error, it is useful to understand not only the physical root cause but also the systemic and behavioural context, why the problem by-passed opportunities for that problem to be captured, before becoming manifest as an error requiring rework.

5.5.2. Costs

Several recent studies have provided support for the use of 5% as an average approximation of the cost of rework in construction. For example, a study in 2009 of 177 construction projects found that the average owner reported rework cost was 5% across all projects. Further, research by the USA’s Construction Industry Institute (CII) reveals that direct costs caused by rework average 5% of total construction costs (CII 2005) (but range from 0-25%). This correlates well with the information that we have received from the steering group. However, these figures relate only to the cost of errors and defects which are recorded. From our research it is apparent that there is a significant further cost of rework which is completed before cost information is captured. The figures relate to the direct cost of errors. From our research it is apparent that there is a significant further indirect cost of errors defects. The implication of this is that the UK construction industry spends substantially more than £5bn per annum in rectifying errors and defects.

5.5.3. Root Causes

The literature identifies the following principal root causes of error:

Client changes.

Design errors by both the client’s design team and the contractor’s design team.

Failures in planning the execution of site works.

Time pressures.

Cost pressures.

The fragmentation of the industry.

5.5.4. Conclusions

Reducing client changes, improving design, and improving project planning are typically identified as the means by which the cost of reworking can be most effectively reduced. The Construction Industry Institute (CII) has presented what appears to be a useful Field Rework Index, a simple 14 question survey that is reported to provide a reasonable indication of the risk of rework on a project.