Improving the Reliability of Roof Constructions: Report From CIB/RILEM International Committee Keith Roberts, CENG, MICE, MIStructE Roberts Consulting 2 Gardiner Close, Abingdon, Oxon, OX14 3YA, UK Phone: 0044 1235 529683 • Fax: 0044 1235 529693 • E-mail: mail@robertsconsulting.co.uk 2 9 t h RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 2 0 – 2 5 , 2 0 1 4 Ro b e r t s • 9 3 Abstract We all hope that the roofs we design, manufacture, build, and approve will perform. In the real world, some roofs do leak and a few blow off in storms. How can we reduce these risks? An international group of roofing specialists has considered the reliability of roofs to identify how to reduce callbacks to completed projects. The presentation includes a case study on the replacement of a computer data center roof where consequential loss following a leak cost $1 million an hour. The International Council for Research and Innovation in Building and Construction/ International Union of Laboratories and Experts in Building and Construction’s (CIB/ RILEM’s) Roofing Materials and Systems Task Group has now published the “Tenets of Reliable Roofing,” which will be presented at this session. Speaker Keith Roberts, CEng, MICE, MIStructE, – Roberts Consulting Keit h Roberts is the principal of his independent firm of consulting engineers. He is a chartered civil and structural engineer who has, since 1991, designed new roofs for commercial buildings and investigated the performance of many different roof constructions throughout the U.K. and Ireland. He has lectured at colleges and written more than 150 technical articles published in the Architects’ Journal, the U.K. roofing magazine, RCI, and the U.S. journal, Interface. He has presented technical papers in North America, including a paper on electronic leak detection at the ISRT Conference held in Washington in 1997. He was chairman of the CIB/RILEM Task Group that examined the reliability of roofing. 9 4 • Ro b e r t s 2 9 t h RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 2 0 – 2 5 , 2 0 1 4 There are concerns about the number of recently completed buildings in which there are recurring rainwater leaks through roofs several years after completion. For some building users, there is a need for a high degree of confidence that the roof will not leak. Such buildings include computer centers, hospitals, civic buildings, and courthouses. This paper refers to the International Council for Research and Innovation in Building and Construction/International Union of Laboratories and Experts in Building and Construction’s (CIB/RILEM’s) Roofing Materials and Systems Task Group that was established at the RCI Convention in Phoenix in 2006 to improve our understanding of the reliability of roofing by seeking improvements from an international perspective. The task group has identified 12 commands or goals. This paper sets out these “tenets of reliable roofing” for the first time to the public. Lessons learned from a reliability engineering approach used for improving vehicle manufacture—such as introducing element redundance—can also benefit the roofing and cladding industry. This paper describes a case study that illustrates how this can be achieved in practice. In reviewing the findings of building inspections, there are often common issues that recur, presenting an opportunity to learn from the experience and to change practice. By developing appropriate means to share feedback in a constructive way, we can improve the reliability of the building envelopes we design and build. 1. Intr oducti on Building owners, their professional advisers, and contractors are becoming increasingly concerned about the number of recently completed buildings with recurring roof rainwater leaks several years after completion. Not all new buildings have roof defects, and it is difficult to build up balanced national pictures. However, the problem does appear to be widespread and occurs on both steep-sloped and low-sloped roofs. For building owners and facilities managers, the reliability of their roofs is important. After completing a project, no one likes to receive a request to return to the site to resolve a problem. For a designer, this could mean lost time with no fee. For a contractor, there are additional costs in time and materials. Delayed payments are a problem for many in the industry, and one of the underlying causes is recurring water leakage, often minor in extent and sometimes due to multiple sources, but enough for a client to justify delaying final payment. For the manufacturer and supplier, there is a potential loss of confidence, making it harder to sell next time. Callbacks are generally unwelcome news, and it would be a good thing to learn from our current experiences and work towards getting the roof right the first time. Some owners need a high degree of confidence that the building envelope will not leak. Such critical buildings include: • Telephone exchanges and Internet server rooms • Offices housing time-dependent operations, trading rooms, and call centers • H ospitals—particularly operating rooms • Civic buildings, including parliamentary buildings and courthouses • Cathedrals and churches • Museums, exhibition halls, and art galleries housing valuable goods • Nuclear facilities • Electrical power facilities The consequences of roof leaks interrupting a building’s operations can result in significant financial losses. Minimizing these risks is in the interest of the building owners and their insurers. This, in turn, should result in better-quality, more robust, and more reliable roofing systems being specified, built, and maintained. Owners expect a leak-free roof for the building’s life, although, unfortunately, the industry isn’t always confident that it can provide such a system. 2. Reliabilit y Studies During the past three decades, there has been a growth in reliability engineering studies, particularly in the aerospace, vehicle production, and electronics industries. Consumers are acutely aware of the problem of less-than-perfect reliability with domestic products such as televisions and vehicles, and they have now come to expect these products to work the first time and continue until they become obsolete. These studies have been highly developed in Japan, where quality and reliability were adopted as national priorities. Owners expect the same levels of service from their building envelopes, particularly meeting the basic requirement of providing a dry internal space. Reliability has been defined as “the probability that an item will perform a required function without failure under stated conditions for a stated period of time” (O’Connor 2002). A crude measure of a roof’s reliability is the number of times the roofing contractor must be called back to the site to resolve a problem. There are many reasons why a roof might leak and result in a need for remedial work. Knowing as far as is practicable the potential causes of failures is fundamental to preventing them, although it is rarely possible to anticipate all of the causes, and a level of uncertainty needs to be taken into account. 3. CIB/RILEM Committee The CIB W83/RILEM Joint Committee on Roofing Materials and Systems consists of 40 roofing specialists drawn from more than 15 countries. In 2006, a task group was established to develop our understanding of the reliability of roofing systems, and Improving the Reliability of Roof Constructions: Report From CIB/RILEM International Committee 2 9 t h RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 2 0 – 2 5 , 2 0 1 4 Ro b e r t s • 9 5 specifically to identify and prioritize practical actions that can deliver improvements. Previously, the committee has examined the topic of roof sustainability and identified key points of best practice, published in the 2001 report, “Towards Sustainable Roofing.” Arising from the discussions, it was recognized that in sustainability audits and other energy studies, the assumption is often made that the roofing system will perform satisfactorily for its full, anticipated life span. Sometimes this is a bold presumption. An environmentally friendly roof system that leaks and needs to be replaced after a couple of years is not a sustainable roof system. The committee has set out to look for examples of reliable roofing practices from different countries, with the objective of reducing the number of callbacks to sites following project completion. 4. Intr oduce Eleme nt Red unda ncy As a starting point, the committee sought examples of roof systems with long lives. A good example is the Shizutani School in Japan, which was commissioned by Lord Mitsumasa Ikeda in 1666. The story is told that his vassal, Nagatada Tsuda, was instructed to build the school and was told that he should construct the roofs so that they would not leak, otherwise he would lose his life. The threat of capital punishment was a keen incentive for the builder to get his roof right the first time! See Figures 1 and 2. The roof system chosen consisted of three layers: clay tiles placed over long wooden plates, which were laid over a shingle roof covering. If the outer tiles were to crack, the water would not get through the roof into the school. The vassal survived, and the three-layer roof system has also withstood the tests of time. Lessons can be drawn from the construction of the Shizutani School roof—lessons which we perhaps have forgotten. The builder recognized three centuries ago the importance of introducing element redundancy, so if one layer is not perfect and does not perform, there is a second layer that can drain the rainwater off the roof. Today, a formal definition of redundancy is “the existence of more than one means for accomplishing a given function.” Good examples of how element redundancy has been introduced into a modern roof construction are tile and slate roof systems with underlayments draining to the eave gutters. In certain weather conditions, some wind-driven rain and snow can find its way below the tiles into the roof space below. The introduction from the 1950s onwards of bituminous underlayments on top of the timber rafters and below tiling battens had the beneficial effect of draining these small amounts of water and snow down to the eave gutters. This development in tiled roof construction more than 50 years ago has improved the reliability of sloped roofs by forming, in effect, a double-layer roof. Since there is redundancy in the system, the passage of rainwater through the outer layer onto the secondary underlayment does not result in failure, because the building remains dry. The roof system is reliable. In the United States, this has been recognized by the Federal Emergency Management Agency (FEMA) in its Design Guide 577 on improving hospital safety in earthquakes, floods, and high winds. Published in June 2007, this recommends that hospitals in hurricane-prone regions install a secondary roof membrane over a concrete deck to avoid water entry if the roof is hit by wind-borne debris. Introducing element redundancy should be encouraged at the design stage of the building envelope. 9 6 • Ro b e r t s 2 9 t h RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 2 0 – 2 5 , 2 0 1 4 Figure 1 – Lecture hall of the Shizutani School, built beginning in 1666. Figure 2 – Roof structure of three layers. 5. Substit ute With Care During the CIB/RILEM committee discussions, it was found that problems have often stemmed from late substitution with alternative products during the construction phase—often to save costs. One view is just to say “no” to substitution, although the designer would still need to provide an explanation. A balancing view is that product substitution can aid innovation, which is important for a developing industry. When drafting project specifications, the designer usually takes considerable care in recommending the chosen products. At the construction stage, the general contractor’s buyers are likely to check that the named products are the most economical, particularly on design-and-build contracts and at the value-engineering stage of major projects. The contractor typically would not be aware of the overall design considerations and how changes could affect other building elements. It is not unusual for designers to come under intense pressure to accept what may appear to be cheaper alternatives. The limited time to make such decisions can result in inadequate assessment. The alternatives may be of a lesser quality, have narrower application limitations, have reduced manufacturer technical support and on-site inspection, and perhaps lack the reassurance of satisfactory references from previous projects. With the benefit of a more rigorous assessment, it may be that product substitution does not give an overall cost saving. Within the committee, members shared their experiences with product substitution. An architectural practice based in the U.S. only would consider substitution at the time of bidding. In Israel, it would be common practice to name three equivalent products named in a specification. In Germany, often only performance requirements would be specified, with product selection left to the bidder. From Israel, a series of criteria for assessing the acceptability of a substitution has been proposed. These include the need to supply a full set of relevant documents, referencing appropriate standards where they exist, restricting the number of requests for changes, and recognizing that it is only the specifying architect or engineer who is authorized to issue an approval that the substituted product is equal to that specified. Product substitution presents challenges to the design and construction team, who could benefit from a formal decisionmaking process that considers performance requirements, cost benefit analysis, service life, and timing. Taking greater care with product substitution and developing an “intelligent caution” approach could help to improve the reliability of the completed roof. 6. C ase S tudy: T he $1 M illi onan- Hour Roof The committee has considered a few case studies of roofs that have not performed. One such case study in which the author was closely involved was on a building in West London that was originally a vehicle repair workshop in the middle of an industrial area. The aged building was refurbished internally with a new mezzanine floor erected to give 150,000 sq. ft. (14,000 m²) of usable floor space. The sloping metal-clad roofs drained to four valley gutters, each running the 260-ft. (80-m) length of the building. In the refurbishment, very little work was carried out to the roof other than lining the gutters. The money was spent on the internal fit-out. See Figure 3. In the summertime in the U.K., it is common to get thunderstorms with very intense downpours of rainfall. One July this happened above the refurbished building, causing the valley gutters to flood, overtop, and leak into the building. The author was called in by the building owners to advise on how to resolve the problem. On arriving at the site, the external building had the appearance of an industrial shed, although inside the building, it was no longer used for repairing cars. The ground floor was now the principal data center for the European operations of one of the London-based merchant banks. Upstairs was a backup trading floor with more than 500 workstations on standby, ready for use. The building owner said the consequential loss from a serious roof leak causing a computer failure would be $1 million an hour. The challenge was to design and install roof improvements that would minimize the future risk of rainwater entering the build- 2 9 t h RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 2 0 – 2 5 , 2 0 1 4 Ro b e r t s • 9 7 Figure 4 – Cross section through valley gutter, as found. Figure 3 – Elevation of building. ing, and that the works on site would need to be carried out without disruption to the operation of the computer data center. See Figure 4. After taking a number of precautionary measures such as installing additional overflows at the ends of the valley gutters, a detail investigation and period of design development was undertaken, identifying ways in which the future risk of overtopping could be minimized. The decision was taken to introduce a second external layer of waterproofing, laid to a slope and draining to the outside of the building. If, in the future, the outer membrane were to be punctured, there would be a secondary internal drainage path down to the original gutters and internal drainage system. This introduced element redundancy. See Figure 5. The new scheme proposed laying prefabricated sandwich panels over the top of the existing pitched roofs, with the membrane’s top face draining into new prefabricated gutters, which in turn had tapered insulation in the gutter to direct rainwater to the gable ends of the building. Of particular concern was the importance of coordinating the details to ease site assembly. The case study also recognized the importance of engaging competent roofing contractors who could be trusted to carry out the works in a methodical manner, following the agreed details and method statements. See Figure 6. Eighteen months after the first call, the building had a new roof with, to the great relief of all parties, no further internal leaks or computer down time. The case study is a good example of how following some basic principles can improve the reliability of a roof system. 7. Lear n From Experie nce Learning from experience is a recommendation that can help prevent problems. Constructive feedback after a project is completed can lead to product development and future innovation. For example, in the U.K., there have been reports of intermittent rainwater leakage through laps in metal panel roof systems laid to shallow falls and particularly on long roof slopes in exposed locations. Site observations have identified sliding movements at end laps in extreme temperatures and the need for fixing and sealant details to be designed to accommodate expansion and contraction. There is an opportunity to learn from these experiences. In the first instance, it should remain the roofing material manufacturer’s responsibility to maintain and update technical information regarding the installation and performance of its roof systems. Independent feedback about building defects has been provided historically by trade associations and government- sponsored research groups such as the Building Research Establishment in the U.K., although this work is diminishing. 9 8 • Ro b e r t s 2 9 t h RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 2 0 – 2 5 , 2 0 1 4 Figure 5 – Cross section through valley gutter, as proposed. Figure 6 – Overcladding work in progress. In Germany, the long-running series of building defects books published by the Fraunhofer Institute in Stuttgart are particularly useful for students and as part of the ongoing continuing professional development for roofing industry members. Technical articles in journals and papers presented at conferences can be an effective means to transfer knowledge. Discussion forums on the Internet are likely to become more common for sharing information, although the need for independent moderation is recognized. By developing appropriate means to share feedback in a constructive way, we can learn from experience and improve the reliability of the roof systems we design and build. 8. Tenets of Reliable Roofi ng Arising from discussions within the committee, a summary was drafted of what appear to be best practices in forming a reliable roof, based on published reports, technical papers, and the experience of the members. It is important that this summary is in a form that will be of practical everyday use for designers, suppliers, and contractors alike, ideally on one page and widely circulated. The underlying basic principles or “tenets” of reliable roofing follow a natural sequence for a typical roofing project, starting with the roof design and contract documentation, proceeding to the procurement of materials and the engagement of contractors, and followed by commissioning and maintenance. It is recognized from the outset that within the industry, we have a significant knowledge base and that on completion of a project, there are opportunities to offer constructive feedback. Considered on its own, each tenet could be considered to be simplistic and no more than common sense. However, when they are considered as a whole, they can make a contribution to promoting good practice in the design, construction, and maintenance of roofing systems. The underlying importance of training and experience is common to all of the tenets. The tenets are given in Figure 7, and the full text with supporting information is given in Appendix A. 9. Concl usi ons The CIB W83/RILEM Committee on Roofing Materials and Systems has examined the concept of roof system reliability and identified a dozen common principles of best practice. The twelve tenets of reliable roofing are a set of commands that promote best practice. These are particularly important for critical buildings such as hospitals, power supplies, and data centers, where there should be adequate resources for the design and construction team. With the translation and widespread circulation of these tenets, the committee hopes that the reliability of the roof systems we design and build will improve. Acknowledgme nts The author gratefully acknowledges the support of the members of the CIB W83/RILEM Joint Committee on Roofing Materials and Systems in contributing to the roof reliability task group. REFERENCES Bauschäden Sammlung (1973-2003), Volumes 1-14, Fraunhofer IRB Verlag, Stuttgart, Germany. “Towards Sustainable Roofing,” CIB Publication 271, July 2001, (cibworld. xs4all.nl/dl/publications/ Publi271.pdf). “Final Report of the Condition Assessment Task Group,” RILEM 166 – RMS/CIB W83 Joint Committee on Roofing Materials and Systems, September 2001. P.D.T. O’Connor, (2002), Practical Reliability Engineering, Fourth Edition, Wiley, Chichester, U.K. K. Roberts, “Reliable Roofing: The Double-Layer Roof,” Proceedings of the International Conference on the Building Envelope and Systems Technology, March 2007, pages 151- 156, Bath, U.K. Design Guide for Improving Hospital Safety in Earthquakes, Floods and High Winds, FEMA 577, June 2007, U.S. H.W. Harrison, P.M. Trotman, and G.K. Saunders, (2009), Roofs and Roofing: Performance, Diagnosis, Maintenance, Repair and Avoidance of Defects, Third Edition, Building Research Establishment Press, Garston, U.K. K. Roberts, “Reliable Composite Roofing: Learning From Experience,” CIB World Congress Book of Abstracts, May 2010, page 25, Salford, U.K. 2 9 t h RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 2 0 – 2 5 , 2 0 1 4 Ro b e r t s • 9 9 Figure 7 – The tenets of reliable roofing. 1. Retain and disseminate knowledge base. 2. Prepare contract documentation. 3. Adopt positive drainage. 4. Introduce element redundancy. 5. Coordinate details. 6. E nsure adequate resources. 7. Substitute with care. 8. Manage effectively throughout the project. 9. E ngage competent applicators. 10. Inspect and test. 11. Plan maintenance. 12. L earn from experience. 1 0 0 • Ro b e r t s 2 9 t h RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 2 0 – 2 5 , 2 0 1 4 1. Retain and disseminate knowledge base Across the world, we already have the knowledge in centers of excellence of how to build a reliable roof. There is an ongoing need for the transfer of this knowledge through training and continuing professional development. The loss of experienced personnel through retirement or redundancy is a threat. 2. Prepare contract documentation The drafting of project-specific drawings and specifications is an essential early stage. There should be proper coordination with other roof-related disciplines, including structural, heating and ventilating, and plumbing. Consideration should be given to the long-term maintenance of the roof, including a safe means of access. 3. Adopt positive drainage Roofs laid sloped to drains have fewer problems and less callbacks to site to resolve leaks after completion. Drainage outlets need to be properly designed with sufficient capacity and connected into the rainwater goods system. 4. Introduce element redundancy A double-layer roof is a good example of a roof that has inbuilt redundancy. If the outer layer were to leak, there would still be a secondary drainage path below so that rainwater does not enter the building. Another practical example is the introduction of an overflow pipe through a roof parapet. 5. Coordinate details A significant number of roof callbacks are a result of flaws in the detailing of the roof, from design to installation. Careful attention should be given to the detailing by adopting standard details that are simple, practical, and robust. On complex roofs and on large roofs with significant repetition, the construction of a full-scale mock-up in advance is helpful for checking that the details can actually be built and for seeking improvements. 6. Ensure adequate resources Resources are needed to build a roof—both in terms of finance to pay for the materials and labor, and time to do the work. At the bidding stage, assess the full-life costing of the roof when comparing alternative schemes and methods. 7. Substitute with care Changing a specified part of a roof system to an alternative product can introduce unforeseen difficulties in the future, such as a lack of compatibility, ultimately increasing the risk of water ingress. Documentation of product performance is needed in the consideration of an alternative. A cautious approach should be adopted to the substitution of products, particularly if the changes are hurriedly proposed, whilst recognizing the value of innovation in construction. 8. Manage effectively throughout the project For the construction of a reliable roof, there is a need for advance planning to ensure the proper sequencing of work, together with the timely supply of the correct materials, tools, and equipment. On completion of a roof, it is important to protect the finished surfaces from damage by follow-on trades. 9. Engage competent applicators Roofing failures are reduced where there are established training schemes for roofing tradesmen, supported by organized roofing trade federations. Applicators should have experience in the chosen roofing system. 10. Inspect and test Checking the quality of materials and work during the roof assembly is recommended, recognizing the difficulties in putting the roof right after completion. Nondestructive test methods can be beneficial in identifying timely corrective actions. 11. Plan maintenance After completion, there will be an ongoing need for the building owner to take responsibility for planned maintenance in the form of routine inspections and good quality repairs as necessary. Restricting roof access to only those who have a need to go onto the roof minimizes everyday damage. 12. Learn from experience On completion of a project, there is an opportunity to critically review the performance and to give constructive feedback to the designers, contractors, and manufacturers. This leads to product development. This virtuous circle can reduce the repetition of mistakes whilst raising general standards and the expectations of the building owner that the roof will be reliable. This is a motivator for innovation. APPENDIX A TENETS OF RELIABLE ROO FING As roof systems become more complex, we face the risk of losing sight of the principal objective of a roof in providing shelter from the wind and rain. As an industry, we all aspire to build roofs that meet the building owner’s need for protection from the weather. It is recognized that on occasion, roofs do leak or become detached. The CIB/RILEM Joint Committee set out to identify specific actions and priorities that will improve confidence. Each command or “tenet” on its own is no more than common sense. However, when read together, the instructions will help to improve the reliability of the roofs we design and build. The need for training and experience is common to all of the tenets. Objective: To collectively identify actions and priorities that can improve the reliability of roofing systems.
