Background of Climate Change
Compared to the average of 1995-2014, the annual mean temperature in Hong Kong in 2081-2100 is projected to increase by about 2.0°C and 3.6°C, respectively, under the intermediate and very high greenhouse gas emissions scenarios (Hong Kong Observatory, 2025a). In the presence of the accompanying seasonal and diurnal variations, the extreme temperatures during summer will be a threat to not only human livelihood, but also various components of infrastructure.
As bridges are susceptible to extreme temperatures, a study has been carried out on the temperature effects on highway structures due to climate change, eventually leading to some revisions to the local design code for bridges. This article highlights some implications of extreme temperatures due to climate change for bridge design.
Effects of Climate Change on Bridge Design
Apart from the higher extreme temperatures due to climate change, the tropical cyclone intensity is also expected to increase with global warming (Hong Kong Observatory, 2025b). Very hot weather is often observed ahead of an approaching tropical cyclone because of the subsiding air at its outer circulation, as confirmed by the maximum temperatures recorded by the Hong Kong Observatory. If the bridges are not capable of accommodating the additional movements and/or internal forces caused by the extreme temperatures due to climate change, any damage so induced could cast doubt on the ability of the bridges to resist the subsequent strong winds due to the approaching tropical cyclone. Therefore, preparation is important.
Evolution of Design Code for Bridges
Up to early 1990s, bridges in Hong Kong were designed to “Chapter 4 Design of Highway Structures” in Volume V of the Civil Engineering Manual published by the then Public Works Department of the Hong Kong Government. Then this chapter was expanded and published separately as the Structures Design Manual for Highways and Railways (SDMHR) in 1993. Up to SDMHR 3rd edition published in August 2006, all these design manuals had referred to the British Standards, various documents of the UK Highways Agency, etc.
Following a major overhaul, the design manual was published as SDMHR 2013 in May 2013, which referred instead to the Eurocodes, the UK National Annexes, etc. Apart from the adoption of a new design philosophy of the Eurocodes, SDMHR 2013 also introduced a more up-to-date methodology for seismic design. Moreover, based on the results of various studies of the projected climate change effects in this region, Revision 5 of SDMHR 2013 was published in December 2023 (Highways Department, 2023b) encompassing the updated requirements in respect of the projected thermal and wind actions due to climate change towards the end of the 21st century.
Thermal Action and Effects of Climate Change
The temperature of a bridge deck changes in response to the changing environment, resulting in the diurnal and seasonal variations. During the daytime, the bridge deck is heated up primarily by solar radiation. During the night-time, the relatively hot bridge deck in summer may also lose heat through outgoing radiation. Within the solid section of the bridge deck, heat transfer takes place by conduction. Apart from the convective heat transfer between the bridge deck and the environment, there may also be similar heat transfer with the interior air void(s) for box girders. The non-uniform temperature distribution over the deck section is often split into the uniform temperature component and the temperature difference for analysis.
Investigations of the temperature distribution of bridges in Hong Kong have begun since the 1980s, covering concrete, steel and composite decks. Typically, they comprised the development of numerical models of heat transfer and calibration with available field data, followed by the prediction of temperature effects for various design scenarios. Examination of various parameters governing the temperature distribution in bridge decks has identified the ambient air temperature as the parameter most affected by climate change, while the seasonal and diurnal variations of solar radiation are expected to be relatively unaffected. The updates to the thermal action in Revision 5 of SDMHR 2013 (Highways Department 2023b) to reflect the potential effects due to climate change have largely been based on an earlier study conducted by the University of Hong Kong. While the temperature difference is relatively unaffected by climate change, the uniform temperature component is substantially affected. Figure 1 compares Table 3.17 in Revision 5 of SDMHR 2013 to that of Revision 4 of SDMHR 2013 in January 2023 (Highways Department 2023a), where the changes are indicated in colour. For example, the maximum uniform temperature component for concrete bridges has increased from 36°C to 45°C, which is quite substantial.
Rational Design Methodology
The rational design methodology of bridges can be explored by examining the structural behaviour of a typical straight multi-span continuous bridge with different support systems under the common design actions possibly with conflicting demands. Under transverse horizontal loading, each pier is normally designed to take up the loading on its own tributary area. However, the response to longitudinal loading will depend critically on the choice of support system, which will be further examined below.
With the higher uniform temperature component due to climate change, an obvious solution is to provide a bearing support system that allows relatively free longitudinal expansion movements. However, this also implies that the horizontal resistance along the bridge will be provided mainly by the support at virtual fixity. Before 2013, this scheme should have been possible as the nominal seismic acceleration was only 0.05g then, where g is the acceleration due to gravity. Even with the partial safety factor of 1.4, the effective seismic acceleration was only 0.07g, as shown in the 5-span continuous bridge in Figure 2. With the adoption of a higher reference peak ground acceleration of 0.12g and the use of dynamic approach with suitable response spectra as prescribed by SDMHR 2013 (Highways Department, 2025), it is desirable to adopt a proper strategy for providing the necessary horizontal resistance.
Table 4.1 of SDMHR 2013 (Figure 3) provides different importance factors to cater for bridges of Importance Classes I to III with the return periods of 475, 1000 and 2500 years, respectively, for determination of the design ground accelerations. While it is imperative to provide the necessary seismic resistance accordingly, the design should not hamper the structural behaviour under other commonly encountered actions, as earthquakes close to the design value may not even occur within the normal design life of 120 years.
The compliance criteria for seismic design include the no-collapse requirement (at the ultimate limit state) and the minimisation of damage (at the serviceability limit state). Eurocode 8 allows a bridge to be designed for essentially elastic, limited ductile or ductile behaviour under the design earthquake. Designing the bridge substructures for essentially elastic seismic behaviour will result in more bulky piers, which may be wasteful. Going for limited ductile or ductile behaviour under the design earthquake will allow the piers to be less bulky, although some damage to the bridge is expected under the design earthquake.
Figure 4 shows the likely outcome of the design scheme for 0.12g with response spectrum analysis (RSA) for elastic behaviour if the horizontal seismic action is taken entirely by a single pier with fixed bearings, and the pier and its foundation must be made much stronger. This is undesirable in respect of both economy and aesthetics. If the two piers adjacent to the central span are made monolithic with the deck, thus sharing the horizontal seismic action, the piers can be made less bulky even for elastic behaviour (Figure 5). Obviously making all four piers monolithic with the deck (Figure 6) enables all of them to share the horizontal seismic action and achieve some economy.
However, both schemes as shown in Figures 5 and 6 are sensitive to the deck expansion movements that will increase due to climate change. Making the piers slenderer will help reduce the internal forces induced by thermal action. If one designs for limited ductile or ductile behaviour, the piers can be made even slenderer as shown in Figures 7 and 8. There are two benefits. On the one hand, the internal forces induced by thermal action can be further reduced. On the other hand, the fundamental period of horizontal vibration will increase, and very likely the spectral acceleration for seismic design will further come down, thus reducing the seismic action.
Therefore, proper seismic design also caters for preparation for the extreme temperature effects due to climate change.
Checking of existing bridges
With the thermal action due to climate change estimated and the design manual updated accordingly, a sensible question to ask is whether the existing bridges designed to the older versions can cope with the higher thermal action towards the end of this century. There is no substitute for detailed checking of each of the bridges. However, most of the bridges of short- to medium-spans in Hong Kong are built of concrete. As most of the shrinkage strain should have taken place within a few years after the completion of construction, the resulting shrinkage movements should have created extra room at the bearings and movement joints for the additional expansion movements in future due to climate change.
It is known that the concrete produced from Hong Kong granite aggregate gives rise to higher shrinkage (Kwan et al., 2010) than that predicted by Eurocode 2 Part 1-1 (BSI, 2004). Hence SDMHR 2013 specifies a modification factor of 2.3 to give the drying shrinkage strain. From the experimental results of Kwan et al. (2010), this multiplicative factor may vary from roughly 1.7 to 2.3 depending on the relative humidity of the environment. For the estimation of loss of prestress and shrinkage movement at bearings and movement joints, it is appropriate to adopt a supremum value of 2.3 as required by SDMHR 2013. However, for the estimation of the beneficial effects of shrinkage in providing some room for additional expansion movement in future due to climate change, the adoption of a lower value of modification factor taking proper account of the relative humidity or even the infimum value may be more appropriate.
References
- BSI (British Standards Institution) (2004). Eurocode 2: Design of Concrete Structures. Part 1-1: General Rules and Rules for Buildings. BSI, London, BS EN 1992-1-1: 2004.
- Highways Department (2023a). Structures Design Manual for Highways and Railways 2013 Edition, Revision 4, January 2023.
- Highways Department (2023b). Structures Design Manual for Highways and Railways 2013 Edition, Revision 5, December 2023.
- Highways Department (2025). Structures Design Manual for Highways and Railways 2013 Edition, Revision 6, March 2025.
- Hong Kong Observatory (2025a). Annual mean temperature projection for Hong Kong. Available at: https://www.hko.gov.hk/en/climate_change/proj_hk_temp_mean.htm (Accessed: 22 April 2025)
- Hong Kong Observatory (2025b). Global climate projections: tropical cyclones. Available at: https://www.hko.gov.hk/en/climate_change/proj_global_tc.htm (Accessed: 22 April 2025)
- Kwan AKH, Au FTK, Wong HHC and Ng PL (2010). Shrinkage of Hong Kong granite aggregate concrete, Magazine of Concrete Research, Volume 62, No. 2, 115-126.
About the Author
Prof. Francis Au
Co-Lead of AIS Academy
Ir. Professor Francis Au brings with him a wealth of over 40 years of experience spanning both the industry and academia. He began his career with Maunsell Consultants Asia, contributing to new town projects, highway initiatives, and marine engineering endeavours from 1981 to 1988. In 1988, he transitioned to an academic post at the University of Hong Kong (HKU). Besides teaching various civil engineering courses, he has been actively engaged in research in bridge engineering and concrete structures until his retirement as Professor and Head of the Department of Civil Engineering at HKU in July 2023. Professor Au continues to be active in research and consultancy work, investigations, and service in various professional institutions.
