Causes of defects associated with tolerances in construction: A case study

18 Defects associated with dimensional and geometric variations (tolerance issues) are amongst 19 the most costly and recurring defects in construction projects, yet the identification and 20 mitigation of the causes of tolerance issues appear to be lacking in the construction industry. 21 To enable the development of widely acceptable solutions for the perennial challenge of tolerance management, a more in-depth understanding of causes of tolerance issues should be 23 established. The aim of the research presented in this paper is to identify the potential causes 24 of tolerance issues in construction based on a literature review and empirical studies. This 25 research uses a case study approach. The empirical data is collected through direct 26 observations, group interviews, semi-structured interviews and document reviews. Having 27 triangulated the findings, a list of eighteen potential causes were derived for the eleven 28 observed tolerance issues in two case study projects. The contribution of this paper to 29 knowledge in engineering management is fourfold: (1) the limitations of prior studies on causes 30 of tolerance issues are revealed, (2) the empirical studies led to not only verifying and refining 31 the causes collected from the literature by considering them in the context of the identified 32 tolerance issues, but also finding new causes in the context of tolerance management when 33 compared to literature, (3) the identified causes provide insight into reasons behind the 34 recurrence of tolerance issues across the industry, and (4) it investigates the causes of tolerance 35 issues while balancing managerial and engineering views. The findings of this study provide 36 a pivotal basis for construction practitioners to develop effective solutions for tolerance 37 management whereby tolerance risks can be identified and mitigated in a prescient manner, 38 which can result in a significant amount of savings. 39

tolerance management, a more in-depth understanding of causes of tolerance issues should be 23 established. The aim of the research presented in this paper is to identify the potential causes 24 of tolerance issues in construction based on a literature review and empirical studies. This 25 research uses a case study approach. The empirical data is collected through direct 26 observations, group interviews, semi-structured interviews and document reviews. Having 27 triangulated the findings, a list of eighteen potential causes were derived for the eleven 28 observed tolerance issues in two case study projects. The contribution of this paper to Defects associated with tolerances, called tolerance issues hereafter, are amongst the most 43 common and recurring defects in construction projects (Landin 2010;Talebi et al. 2020a). 44 Tolerances issues (e.g. lack of fit, misalignment between components, aesthetically The findings and contributions to knowledge are discussed. Finally, conclusions drawn from 70 the research and areas for further research are presented.

71
Causes of tolerance issues in construction 72 It is important to first investigate causes of tolerance issues in the literature. Berg Milberg (2006) argues that (1) incomplete or missing tolerance information in specifications, 80 which results in multiple interpretations of tolerance requirements by contractors and 81 inspectors, (2) incompatibility between the specified tolerances and process capabilities, and 82 (3) poor datum selection, being the main causes. Process capability in the context of tolerance 83 management, is defined as "the likelihood that a process…will result in an outcome that meets 84 a given tolerance specification. It may be represented by a probability distribution to describe 85 the variation in the geometry of the material output of a process under normal operating 86 conditions" (Tommelein and Ballard 2018). A datum is a theoretically exact geometry from 87 which the geometric or location characteristics of a feature are established (Sun and Gao 2018). 88 The latter two causes show that tolerance management in construction has its roots in the 89 manufacturing industry.  During construction, building structures are not protected but are exposed to environmental 138 conditions; particularly changes of temperature, rain, snow and humidity, sometimes for a long 139 time. Such environmental conditions may lead to changes in form and size of components after 140 they are constructed and subsequently result in tolerance issues (Alexander 2014). This is 141 evidently seen in bridge construction and in large buildings where expansion joints are 142 necessary for ensuring that a structure is not over-constrained due to thermal changes. Changes in temperature, including temperature difference across a component or changes in the average 144 temperature of an assembly, are a source of building movement, especially in steel structures.

145
Rain and snow can directly damage the accuracy of the final work (e.g. flatness of fresh 146 concrete on metal decking in composite construction) (Alexander and Lawson 1981)  This research adopts a case study approach. Such an approach is suitable when describing, 175 explaining, and exploring a contemporary phenomenon, and gaining an in-depth understanding 176 of real-world events (Yin 2013). In other words, the purpose of the case study is not to simply 177 describe the events within a real-world context but rather to investigate underlying reasons as 178 to why and how those events actually occur (Eisenhardt 1989   Rationale for case selection 190 No widely accepted strategy exists for the selection of right cases beyond the advice to select 191 cases which are 'most likely' to address the research aim (Brinkmann 2013). Accordingly, purposive sampling was adopted for the selection of cases and participants in the interviews.

193
The purposive sampling highlights the importance of conscious decision-making and is used were also identified during the empirical studies, that is, the selected cases were informative 203 and the purposive sampling in this study ensured that appropriate cases have been selected.

204
Details of the case projects and their development stages can be found in Table 2.   Table 3.  Table 3. Details of the tolerance issues identified in projects A and B.

215
Two group interviews were then conducted to validate and refine the description of the 216 tolerance issues. The details of the participants in each group interview are given in Table 4.

The managing directors of the general contractors in projects A and B suggested interviewees
218 based on the following criteria: (1) all of them had more than 10 years of experience in dealing 219 with tolerance issues, (2) they were involved in the project from the beginning, and (3) were 220 fully aware of the identified tolerance issues in each project. This ensured that purposive 221 sampling had been followed carefully and the right participants had been invited to those group 222 interviews, which is more important than the number of participants (Brinkmann 2013).

230
Of note is that the literature review in this research was carried out in two stages. The first 231 stage was completed before starting the semi-structured interviews in 2016, and led to a list of 232 causes of tolerance issues. The list was sent to the participants of face-to-face interviews for 233 their reference. In particular, the participants were asked to consider unique characteristics of 234 tolerance issues identified in their own project. The second stage of literature review continued, 235 and it covered the most recent and pertinent literature. However, as can be seen from Table 1      Check Sheets included information about the permissible deviations of the plumbness of the 332 SFS studs" (Interviewee 1), or "how and when they should be measured" (Interviewee 1). As 333 a result, the tolerance problem with the SFS studs was recognised after they were handed 334 over by the cladding subcontractor (Interviewee 7), when "they started to build their system 335 on the SFS studs" (Interviewee 7). Thus, this problem was due to ineffective quality control 336 documents. Also, Interviewee 4 says that given "the operatives had to complete their work as 337 quickly as possible, they were not that much concerned about the quality of their work".

338
Hence, poor workmanship was another cause behind this tolerance issue.

339
The cladding subcontractor developed a design in which the offset from the steelwork to the 341 face of the stone panels was 272 mm. In this case, the cladding system could absorb 32 mm 342 of deviations due to the inclination of steel columns and stone panels. The Architect later 343 increased the offset to 290 mm (Interviewees 1, 2, 6, 8) (Figure 4). This was to accommodate 344 the installation between the steelwork and cladding. "There was miscommunication between 345 the architect and the cladding subcontractor about the required distance from the steel to the 346 face of the stone panels" (Interviewee 2). The subcontractor's input was delayed 347 (Interviewees 1), and "the architect was not convinced to change the design" (Interviewee 1) 348 as "the connection type between the steelwork and cladding system had been designed"   When the cladding subcontractor put the stone panels on and the dead load was applied to the 361 steel frame, the stone panels started to sag. This meant that the cladding did not stay at the 362 correct level and, in general, everything was sinking downwards. It was noticeable that the 363 gap between the channel and the stone panel in some areas were bigger and the gap was not 364 consistent all the way through ( Figure 5). There was miscommunication between the steelwork contractor, cladding contractor, and the general contractor did not perceive the 366 importance of having movement joints, thus, denied to have it (Interviewees 1, 7, 8, 9). In the 367 specification for the cladding system prepared by a consultant, it was stated that "movement 368 joints are not required". However, it turned out that this tolerance-related information in the 369 specification was inaccurate. The Interviewees 1 and 9 believed that using movement joints    is essential to ensure that the electrically operated shutter doors will fit in the doorways (i.e. (Interviewees 10, 14). "There was no 394 communication of tolerance information before construction whatsoever" (Interviewee 11).

395
The only information communicated between the site manager and the steel subcontractor  (Interviewees 10, 12, 14). Moreover, the structural designer had tried to use steel and 402 ancillaries as little as possible (Interviewee 13). The two sides of the doorways were "neither 403 connected to each other nor [were] they … fixed to the ground" (Interviewee 11), that is, they 404 are free standing. The inconsistency between a tolerance requirement (i.e. fitting the shutter 405 door without any rework) and its budget led to a situation that by aligning one side of a 406 cladding rail, the other side was becoming out of alignment (Interviewees 10-12) (Figure 7).  The frames for personnel doors were neither plumb nor square. The personnel doors were 456 'squeezed' too tight, resulting in a non-square condition, and they could not be shut and 457 opened properly. According to the Interviewees 13 and 15, the cladding subcontractor had 458 measured the distance between the posts making the doorframes at the bottom and top using a conventional measuring tape. However, "they ignored the fact that the doorframes actually 460 can be oriented either to left or right side even though when those distances are the same" 461 (Interviewee 14) and a measuring tape is an inappropriate measurement instrument to check 462 the plumbness tolerance of the doorframes (Interviewees 11, 13).  The findings from the observations and interviews helped to verify and refine the causes 481 collected from the literature by bringing them in the context of the identified tolerance issues, 482 and also to find causes that had not been considered in the tolerance management body of knowledge. 'Insufficient tolerance information in specifications' is a cause for tolerance issues 484 identified in the literature. This cause was refined further during this study as it was 485 demonstrated that inaccurate and fragmented tolerance information in specifications led to 486 tolerance issues 6 and 7. Accordingly, it was suggested that the 'insufficient, inaccurate and 487 fragmented tolerance information in specifications' is the correct cause (Interviewees 1, 2, 4,  The existing scattered attempts within academia to tackle specific causes of tolerance issues 530 found in each research work, rather than devising a solution to treat all of them, will not bring 531 about a panacea for tolerance management. Existing solutions devised to proactively prevent 532 these issues must be like putting together pieces of a puzzle, rather than creating isolated solutions to remedy a number of specific causes. This is because all the identified causes bear

538
The list of causes in this paper helps to develop a solution by which the identified risks and 539 their causes can be tackled and also provides insights to practitioners about potential tolerance 540 risks and their causes that need to be mitigated in projects. That is, knowing the causes will 541 improve the competency in the industry to deal with tolerance risks in a prescient manner, 542 which is a core principle of an effective tolerance management practice (Talebi et al. 2020a).

543
Therefore, this research is expected to provide a pivotal basis for developing a solution for 544 tolerance management whereby the identified causes can be tackled and a significant amount 545 of savings can be made.

547
The aim of this research was to identify potential causes of tolerance issues in construction.

548
The literature was used as a basis to create the preliminary list of causes. The case study 549 approach adopted in this research allowed to collect empirical data through direct observations 550 in two case projects, two group interviews, sixteen semi-structured interviews and document 551 reviews. Direct observations and group interviews were to identify examples of tolerance issues 552 in practice. Semi-structured interviews helped to collect rich data from which the experience 553 of participants about causes of the observed tolerance issues could be captured. In other words, 554 the semi-structured interviews were conducted to examine the causes of tolerance issues 555 encountered in real-world to avoid the subjectivity of reflecting the authors' or practitioners' 556 perceptions only. The document review was to corroborate evidence collected from the interviews and it was also used for verifying information about the case projects that had been 558 presented in the interviews.

559
The findings from the literature and empirical studies were used to generate a list of eighteen 560 causes for the observed tolerance issues in the two case study projects. This list is also expected 561 to give an insight into causes behind the reoccurrence of tolerance issues in other projects  This paper also consolidates scattered insights on causes of tolerance issues into a refined 577 and unified list of causes. The findings presented in this paper are expected to be a starting 578 point when identifying causes of tolerance issues and developing solutions for tolerance 579 management. Of note is that the construction industry is argued to lack a widely accepted 580 solution for tolerance management due to the lack of in-depth understanding of causes behind 581 the recurrence of tolerance issues.
Three limitations of this case study research should be explained. First, the empirical studies 583 revolve around commercial and industrial buildings and future research may find more causes 584 applicable to other types of construction projects. Second, this study was conducted in the UK 585 and may be affected by the special characteristics of the construction industry in this country. 586 However, it is arguable that these two limitations are partially mitigated as a result of reviewing 587 the literature on tolerance issues in various types of construction projects conducted in different 588 countries. Third, the causes identified for each tolerance issue were only inferred from the 589 accounts given by the interviewees, that is, other causes for each tolerance issue may be 590 conceivable. For example, a closer look at the tolerance issues and their causes reveals that 591 'ineffective quality control process' and 'poor workmanship' could be attributed to tolerance 592 issue 8, 'poor communication of tolerance information' could be attributed to tolerance issue 593 10, and 'insufficient, inaccurate and fragmented tolerance information in specifications' could 594 be attributed to tolerance issue 11. It can be argued that this limitation arising from the 595 characteristics of the case study research was mitigated by sharing the potential causes of 596 tolerance issues identified from the literature with interviewees in order to increase their 597 awareness of the topic. Future research may attempt to undertake a similar study to find causes 598 of tolerance issues in other projects in order to generalise the causes, that is, developing a list 599 of causes without considering it in the context of a specific tolerance issue and project.

601
Some or all data, models, or code that support the findings of this study are available from 602 the corresponding author upon reasonable request.