HomeContact us Gallery Site Map Resource Centre Search
About us People News Publications Education & Training Events
Publications | Education and Training | Current & Recent Projects | Events | Research Opportunities
 

Download
Technical Paper 1(1,82MB PDF)



Cover Page

The Lack of Preparation Problem

The Flash Flood Problem

The Landslide Problem

The Rising Groundwater Problem

The Insurance Claims Handling Problem

The Flood Claims Excess Problem

The Buildings Regulations Problem

The Dams and Reservoirs Problem

The Canals and Waterways Problem

The Water Framework Directive Problem

The ODPM versus Defra Problem

The Priority Problem

The Participation Problem

Next Section

Technical Paper 1
Flood Risk & Insurance in England and Wales: Are there lessons to be learned from Scotland? - David Crichton


The Dams and Reservoirs Problem

Disaster in Dolgarrog

On Monday, 2nd November, 1925 , there was a blow out of the lower section of a portion of the Eigiau dam in the Conwy valley in North Walesi . The water scoured a channel 70 feet wide and 10 feet deep, as 50 million cubic feet of water surged down to the Coedty reservoir below. Coedty reservoir was nearly full at the time and the spillway had to cope with a surplus discharge well in excess of its designed capacity. The dam was overtopped, washing away the embankment, and the core collapsed. There was an almost instant release of 70 million gallons of water. A wall of water, mud, rock and concrete hit the village of Dolgarrog at 9.15pm . Fortunately many of the villagers were attending a film show at the village Assembly Hall out of the path of the flood, and 200 workers were working late in the nearby aluminium factory, otherwise more lives would have been lost. As it was, ten adults and six children were killed and many houses were destroyed. At the subsequent inquest, the deputy coroner said that "as Dolgarrog had a floating population, . it was impossible to estimate the extent of the catastrophe in regard to human life". A rather unfortunate turn of phrase, especially as one body was not found until ten months later, having been carried away down the River Conwy.

Huge boulders, the size of houses, can still be seen in the village.

It later transpired that the general manager and board of directors of the company which owned the dams knew that there were defects in them from the beginning, but chose to keep the facts secret. No one was ever held to account, and two of the streets in the rebuilt village were named after directors of the company ii .

(i) Thomas, D., W., 1997 "Hydro Electricity in North West Wales " National Power plc, Llanrwst, Wales.

(ii) Draper, C., 2002 "Walks in the Conwy Valley " Gwasg Carreg Gwalch, Llanrwst.

 

It should be emphasised that no lives have been lost in the UK from dam failure since the Dolgarrog disaster in 1925, however failures do occur around the world. In 1959 when the Malpasset dam in France failed, 421 people died, and in 1963, overtopping of the Vaiont dam in Italy caused by a landslide resulted in 1,189 deaths, even though the dam itself remained intact. In 1972, a dam in West Virginia , USA failed causing 125 deaths. In 1976, the Teton dam in Idaho , USA , failed during its initial filling, killing at least 11 people. The worst dam disaster was the failure of the Banqiao Dam which collapsed in 1975 during a typhoon. 85,000 people were killed, and a further 145,000 died from the subsequent epidemics and famine. (30,000 reservoirs in China have serious safety problems, according to the Chinese government. Between 1954 and 2003, 3,484 dams collapsed in China .)

Dam failures are increasing:

•  In May, 2004, a coffer dam collapsed in China , killing 14. It had been built by the same company which is building the Three Gorges Dams.

•  In June 2004, after heavy rain, the Camara Dam in Brazil burst. It killed five people and left 3,200 homeless. It had only been completed in 2002.

•  In July, 2004, heavy rainfall destroyed 13 small dams in New Jersey in the USA . 19 dams have burst in New Jersey alone since 1999. (According to the US Association of State Dam Safety Officials, the US needs to spend $10billion on the most dangerous dams.)

There still seems to be a defensive and complacent attitude amongst dam engineers in Britain who point out that modern British dams and reservoirs are designed and built to very high standards. This may well be true, but most dams in Britain are over 100 years old and built of earth.

In the future, the safety margins will increasingly be eroded by climate change. So far as current safety standards are concerned the main concern is the secrecy surrounding the condition of dams and raised embankments. There are also concerns about the secrecy surrounding dambreak inundation maps, and the lack of preparedness of the emergency services in dealing with a catastrophic failure. The author was a member of a steering committee for a detailed report on reservoir risks produced by CIRIA, in 2000, and had access to a great deal of information which is not in the public domain. While the author is not at liberty to disclose such information, the following public domain information should be enough to illustrate the scale of the issue.

The Reservoirs Act, 1975, applies to all reservoirs holding or capable of holding more than 25,000 cubic metres of water. There are over 2,500 such reservoirs in the UK of which 530 are large enough to be included in the World Register of Large Dams. Owners of dams covered by the Act are obliged by law to have them inspected every ten years by a civil engineer from a special panel, but the law does not specify the details of the inspection nor that the results should be published. In practice, the thoroughness of the inspection depends almost entirely on how much the dam owner is prepared to pay, and the author is not aware of any case where the results have been published, or even given to local authority emergency planning officers or the emergency services. (The author would welcome details from anyone who knows of such a case.)

Dam owners also refuse to issue dam break flood inundation maps. This could mean (and indeed this has happened) that planning officers for the local authority might grant planning permission for new housing developments within the area which would be flooded if the dam failed, simply because they did not know that the area was within the danger zone.

By contrast, in France , everyone living in areas at risk from dam break is fully aware of the fact, and these areas are subject to frequent evacuation exercises. Informal comments from engineers would seem to indicate that they believe the British are more likely than the French to panic if they were given such information.

In Sweden , detailed flood maps which include dam break inundation maps are readily available to the public. Why not in Britain ?

Climate change could well lead to an increased risk of failure of British dams, some of which are more than 200 years old. Failure can be caused by many factors, - for example climate change could lead to subsidence of the dam foundations, landslip into the reservoir, or overtopping due to heavy rainfall. Around half of the 2,500 large UK dams have earth embankments, most of them constructed before heavy soil compaction equipment was available. Little is known about the content of such embankments, especially the core, or the extent of internal settlement or disturbance, for example from rabbit burrows.

Droughts could lead to cracking of the embankment wall, and climate change will lead to more droughts in the summer, followed by more rain in the autumn. This could impose additional loads, which were not considered when the reservoir was planned. There could also be additional loadings from increased snowfall in upland areas, followed by rapid snowmelt due to rainfall. Higher windspeeds over the reservoir surface could cause more frequent overtopping, leading to erosion of earth embankments unless suitably protected.

Other possible causes of failure include vandalism of valves, pipe work or controls, terrorism, or aircraft crash.

Many dams are in or near urban areas, for example there is a large reservoir in Brent in London which is very close to housing and aircraft flight paths.

Most UK dams are over 100 years old. A detailed record is kept of defects in dams but this is not published. The reasons for the secrecy surrounding the condition of the nation's dams are not clear, but prudent underwriters are always inclined to assume the worst when information is withheld. The record of dam safety in the UK has been excellent since 1925, but climate change fears might cause some underwriters to reassess the situation. It would seem quite likely that there are people living and working within the danger zone of large dams in the UK . In the USA where information is more readily available, it is known that there are more than 2,000 communities that have been identified as being at risk from dams which are believed to be unsafe.

Growing population and wealth, especially in the south east of England , will lead to greater demand for water, while supply will be reduced by summer droughts. Demand management controls such as water meters can only have a limited effect and groundwater abstraction is near its limit in some areas.

(In other areas, such as Liverpool and London , abstraction of groundwater has ceased, and rising groundwater is becoming a problem - see "The Groundwater Problem" section.)

It is therefore likely that more dams will need to be built, and in Southeast England , these are likely to be near urban areas.

Meanwhile, in 2000, Government introduced a research programme for dams, with some of the survey work sub contracted to the Transport Research Laboratory because of their expertise in checking earth embankments.

One cost effective way to monitor the condition of dams and raised reservoirs is to use "Synthetic Aperture Radar" (SAR) instruments on satellites, using a technique called "PSInSAR" (Permanent Scatterer SAR Interferometry), which can detect sub millimetre movements in the ground or structures. All that is needed is to install special "corner reflectors" on the walls or embankments, and then monitor them for movement. Once a system is established, monthly reports can be produced at relatively little cost. The corner reflectors cost around £1,000 each, but a new technique using SAR transponders costing around £100 each is showing great potential

Italy is so impressed with the technology that it has already installed large numbers of SAR transponders and reflectors. Thames Water has installed seven satellite "reflectors" on the embankments of the Wraysbury Reservoir in West London to detect and monitor sideways or vertical movements using regular satellite images. (RADARSAT passes over every 24 days, while ENVISAT passes over every 35 days.) This technique should ideally be used on all dams and reservoir embankments, not to mention flood defences and bridges.

This new technology could give early warnings of possible breaches in dams, embankments and flood defences.

More dams are likely to be built because of climate change, either retention dams as part of flood alleviation measures, or reservoirs for hydro power or water supply in response to the moves away from fossil fuel power generation and increasing summer droughts.

While dam break inundation maps are confidential, the software used to construct such maps is not. It would be quite possible for insurers to create their own dam break maps and take action to reduce their exposure in areas subject to this hazard. This could become easier after the completion of a major European research project under the EU Fifth Framework Programme called "IMPACT" (Investigation of Extreme Flood Processes and Uncertainty) which is being co-ordinated by HR Wallingford. The project is costing €2.3m and as it is publicly funded, the results are expected to be disseminated publicly when the project is completed. It is examining breach formation, flood propagation, sediment movement and process uncertainty. Early results are extremely interesting, but there is a great deal of work still to do. One of the aims is to produce more user-friendly software for producing inundation maps for dam break throughout Europe . This would make it easier for insurers, local Councils, and the Health and Safety Executive to produce their own inundation maps, and this might encourage the government to end the long standing but unjustified exemption which dam owners have from the requirements of the Control of Major Accident Hazard Regulations 1999 (COMAH).

Meantime, so long as dam condition reports and inundation maps remain secret, prudent insurers may be increasingly likely to assume that properties near to reservoirs may be at risk of flooding from a breach, particularly in the case of older earth embankments, or concrete dams constructed more than fifty years ago.

There are also general environmental problems relating to dams, because although they are a source of renewable energy, they can actually increase greenhouse gas emissions and lead to polluted rivers.

Case Study

(This is an actual case, which is in many ways typical, but the author has omitted details that might identify the dam in question, in order to protect the engineers who provided the information.)

This is a large reservoir originally formed by an earth dam constructed around 150 years ago. The dam wall was raised in the 1970s, on top of the old earth wall, which now forms part of the upstream toe. The reconstruction involved a 330m long, 9m high dam covered in rock armour on both sides to the height of the spillway, with a 1 m wave wall.

It is subject to the Reservoirs Act 1975 as a "Category A" dam, and therefore must be independently inspected every ten years. The last inspection was in 2001, and it is not due to be inspected again until 2011.

The seismic category is "C" (low risk). The latest panel engineer's inspection report makes no attempt to predict what the condition of the dam will be in ten years time, but reports some deterioration and cracking at the time of the inspection. No attempt was made to determine the condition of the core. There is no instrumentation to give advance warning of failure, and no warning or evacuation procedure. The only recommendation from the engineer was to repeat the 1991 recommendation (which had not been actioned) that a security fence be erected to prevent access by children to the dam.

A formal dam break inundation map has not been produced, but a risk assessment report produced in 2002 contains an inundation map showing that 63 recently built domestic properties are at risk. Local planners had allowed these houses to be built because they were unaware of the hazards, and there are no contingency plans for evacuation. The report suggests that a major breach or failure of the dam could produce a flow of 7257 cubic metres a second, with the potential for 338 deaths, owing to these recently built properties, making this a risk "Level 1" dam, in accordance with the CIRIA reservoir risks manual i . (The author was a member of the steering committee for the team that produced this manual).

The spillway leads to a double culvert under a major trunk road that leads downhill to a large town. The diameter of the culverts is 1.2m and they are only inspected annually. There are no trash guards. If the reservoir were to be full, a severe rainfall event could cause a significant flow over the spillway and the trunk road could become a "river" leading flood waters into the centre of a housing area nearby.

The hills surrounding the reservoir are covered in a layer of peat, overlying moraine and rock. The inspecting engineer did not mention (and perhaps is unaware) that peat on a sloping surface is very vulnerable to sliding if a long dry summer is followed by a wet winter, which could cause the peat layer to become top heavy. A significant peat slide into the reservoir could lead to sudden and catastrophic overtopping. There is no monitoring of movement of this peat layer, and the dam owner was unaware of the PS InSAR ii satellite monitoring systems that could give early warning of such an event. The author is currently attempting to persuade the dam owner to commission research into the peat slide risk from a peat expert, and to install satellite transponders.
Dry summers and wet winters are predicted to become more common with climate change, according to the UK Climate Impacts Programme projections iii .

(i) Hughes,A; Hewlett, H W M; Samuels, P G; Morris, M; Sayers, P; Moffat, I; Harding, A; Tedd, P. 2000 "Risk Management for UK Reservoirs." Construction Industry Research and Information Association (CIRIA) Research project report C542. London .

(ii) Permanent Scatterer Synthetic Aperture Radar Interferometry

(iii) Hulme, M., Turnpenny, J., Jenkins, G., 2002. "Climate change scenarios for the United Kingdom . The UKCIP 02 Briefing Report." Tyndall Centre for Climate Change Research, University of East Anglia , Norwich , UK

Climate change could also threaten the safety of this reservoir in other ways. A recent report for government suggests safety margins could be reduced by some 20% with climate change iv .

Council planning and emergency planning officials are not aware of the government's report on climate change implications for dams, and have not been given access to the panel engineer's report, the risk assessment report, or the inundation map v . The local Council recently allowed three sheltered housing complexes to be built in the area and, until the author intervened, was looking at proposals to build a new hospital and primary school there.

(iv) Babtie Group and the Centre for Ecology and Hydrology, 2002. " Climate Change Impacts on the Safety of British Reservoirs " Report commissioned by the Department of the Environment, Transport and the Regions (DETR) now DEFRA, through their reservoir safety research programme. (Unpublished).

(v) Dam condition reports are confidential documents, as are dambreak inundation maps. The Dam Owners Society forbids its members from publishing them in case it causes public concern.



«back to top«