Solar effects on a partially buried reinforced concrete service reservoir

The method of insulating partially buried service reservoirs against solar effects is generally recognised as insufficient to prevent differential movement between the roof and the walls. It is, therefore, common practice to design partially buried reservoirs with a sliding roof/wall joint detail to negate the forces and moments induced by the thermal movements. Currently, the level of additional moments generated in a monolithic joint from solar radiation in this situation can only be estimated as there is a lack of real data. However, Arup’s commission for the design of Cropton Service Reservoir for Yorkshire Water provided the opportunity to undertake a monolithic design. The driver for the use of monolithic roof-wall joints was Yorkshire Water’s aversion to sliding joints principally in view of the required maintenance of the joint but also the potential source the joint provides for bacteriological failure. The School of Civil Engineering, University of Leeds, was engaged to validate certain design assumptions made by Arup. This was achieved by acquiring: 1) long-term thermal data, particularly thermal gradients across the structure and within the individual elements of the structure; 2) strain data for a real full-scale structure with monolithic construction. The decision to instrument the structure came very late into the programme of work for Cropton, i.e. only 3 months before construction was due to commence. The structure had therefore already been designed before it was decided to instrument it. Only a preliminary investigation could be organised to confirm the assumptions made by Arup. The investigation also provided the opportunity to examine the thickness of the gravel layer and the effect that this has on the attenuation of solar radiation and the subsequent temperature differential within the structure. The research validated the initial assumption by Arup of a 10ºC maximum temperature differential between the roof and the walls. It also showed that the temperature within the tank with 175mm of insulation on the roof did not exceed 14ºC. This is the maximum temperature within the tank identified by Yorkshire Water before bacteriological failure may occur. Interestingly, the investigation has also highlighted, year-on-year, a permanent expansion in the roof which is producing additional moments in the walls. These moments exceed those present purely due to actual thermal loading. This year-on-year expansion or ‘ratcheting’ effect is still continuing after almost 4 years and is thought to be the result of swelling and creep/microcracking resulting from the thermal loading. The full extent of this effect has yet to be determined and so it is still under investigation.


This paper describes the monolithic design, the field research that was undertaken to confirm the structural behaviour, and the assumptions made in the design. An indication of the likely cost savings is also made. Parts of this paper was presented at an evening me