Report of the presentation 'Geotechnical aspects of underground storage caverns for oil and gas'by Professor Ulf Lindblom at the annual joint meeting of the British Geotechnical Association and the Geological Society, held at Burlington House on 20 February 2001, by R J Pine, Camborne School of Mines, University of Exeter.
Introduction Professor Ulf Lindblom of Gothenburg was introduced by deputy chairman of the BGA, Professor Barry Clarke, who noted the speaker's background in civil engineering geotechnics and engineering geology, specifically the design of storage caverns worldwide.
Lindblom was the originator of Roda Sten underground rock laboratory in Gothenburg, which he has led since 1989.
This has been used for major studies, on a pilot scale, of both the cryogenic storage of natural gas and the pressurised storage of natural gas with water curtain control of groundwater.
He was also the project manager for research and development in geothermal energy (hot dry rock) in Sweden, from 1986 to 1988, and has been both a consultant in industry and an academic at Chalmers Technical University, Gothenburg.
Oil storage Lindblom explained that the early days of underground storage of oil were during the Second World War when the key issue was security. In this repect underground caverns have obvious advantages over surface tankage. Other early underground caverns were built for hydroelectric power plants.These were originally of modest size, limited by the low productivity of handheld rock drills and small scale mucking equipment.However, oil storage in caverns was developed through the 1950s in Scandinavia, principally using the Sentab system.These caverns were similar in shape to surface oil tanks and were also lined with steel.
It was soon recognised that a lining was not necessary for oil storage. One of the earliest stores, also in the 1950s, involved the use of an abandoned iron ore mine.The mine was naturally flooded and it was shown that by filling suitable parts of the mine with oil, floating on the water, the head of water outside the excavations would exceed the head of oil within and there would be a net inflow of water. In this way oil would be prevented from leaking with no consequent safety and environmental problems.This storage was in use for 30 years.A cross-section of the mine is shown in Figure 1.
One early enthusiast was a Mr Edholm, who was a director of the port of Stockholm.He stored gasoline in a pilot underground storage cavern he had built under his own extensive property in Stockholm, for three years before showing the authorities.He was able to demonstrate theoretically and practically that the balance of water and gasoline pressures provided secure storage without a lining.
In the 1960s there were rapid increases in both size and numbers of caverns, with the development of modern multiboom drilling jumbos and associated excavation equipment.
The modern caverns are large and dramatic.Typical dimensions are 25m wide and 30m to 40m high, excavated by top heading and benching down.The cross-section and length are similar to those of Notre Dame cathedral. Oil is stored on a water bed, which is pumped out as dictated by the water inflow. It is important to avoid or treat high permeability zones to minimise inflow because of the economics of excessive pumping.The caverns are usually 30m to 50m below ground, where sufficient groundwater pressure is found in the hard rock conditions in Scandinavia.Construction access is by means of a ramp and the system is completed with a shaft access, which houses pumps, pipes and seals.
Figure 2 shows a typical oil cavern complex layout.Figure 3 shows the interior of a cavern.The boat is on the water bed.
Gas storage By the 1970s this method of oil storage was commonplace and several hundred oil storage caverns were constructed.
Additional strategic (large) caverns were built as a result of the decade's oil crises.
In terms of technical challenges, the focus turned to the storage of liquefied petroleum gas (LPG), initially as a liquid at about 100C under an over pressure of 3 to 8 bars (30m to 80m head of water).This was achieved successfully with shallow caverns using artificial water curtains to contain the gas.One of the first European examples was at Killingholme in the UK. Earlier about 70 deep caverns for LPG storage were constructed below aquiferous rocks in the US.
Water curtains have their roots in mining.When most mine systems were pneumatic there was a need for high-pressure storage and mined openings were frequently used. An upper drift was filled with water, providing a head to a lower drift into which air was pumped under pressure.
It was recognised that if the water level in the upper drift was too low and the air was maintained under pressure, there was significant air leakage through the rock mass. If the water level was kept high the combined effects of the resulting saturation and elevated pore pressures was much more effective in limiting air leakage. Modern water curtains consist of a grillage of tunnels and boreholes connected by pipework to a header reservoir, usually at ground surface.The water curtain tunnels have bulkheads to allow for substantial water overpressures.
The next logical step was storage of LPG in low temperature caverns at a temperature of about -400C.Shipment of LPG is by refrigerated vessels and it therefore arrives in the cooled liquid state. A combination of design and trial and error revealed that it is quite feasible to store LPG at this temperature, using the resulting ice wall as containment, with the liquid at near atmospheric pressure. In this case water curtains are either not needed or much simplified.
There are now hundreds of LPG caverns worldwide.
Cryogenic storage of liquefied natural gas (LNG) is the next challenge, requiring cooling to -1640C for storage at near atmospheric pressure.This is an extremely challenging temperature because of the consequent shrinkage of the surrounding rock mass. There have been two failed attempts in medium to hard rock in Hackensaw, New Jersey and in Boston, Massachussets; and a successful research attempt on pilot scale in the Boom clay in Belgium. Generally it is viewed as too difficult for current methods.One possibility is to duplicate what now happens above ground using steel tankage and insulation.The rock mass provides further insulation, particularly in comparison with warm weather above ground.
The Roda Sten experimental caverns in Gothenburg have been used successfully to investigate fundamental aspects of both water curtain and cryogenic technologies applied to underground storage.The hydrogeological, geomechanical and thermal bulk properties of the rock mass have been calibrated/confirmed under a number of different configurations.The effectiveness of bulkhead and grouting systems has also been investigated with subsequent operating differential pressures of up to 100 bar.These fundamental parameters have been fed into many subsequent designs.
There is not much activity in Europe on oil and gas storage as there is almost enough capacity.A possible growth area is caverns for diurnal storage of natural gas (not liquefied).The lead has now passed to the Far East, particularly Korea.
Additional examples of the use of cavern storage Leaving aside underground tunnels and caverns for transport infrastructure, other uses of storage are varied and ingenious.A seminal conference was Rockstore 1977 where these were widely publicised for the first time.
Stockholm alone has more than 80km of tunnels that are unrelated to transport infrastructure. Applications include fresh water, electricity, district heating and telephone supply.
Pumped hydroelectric power plants such as Dinorwic in Wales can also be created with underground reservoirs, particularly the lower reservoir.An example is shown in Figure 4 (overleaf ).
Compressed air storage has been used initially for mining pneumatic systems, then for wind tunnels and in recent years with gas turbines for electrical power generation known as compressed air energy storage, (CAES).
In a typical gas turbine, two thirds of the gross mechanical energy from the turbine output is needed to compress the air used in the combustion.By decoupling the turbine and compressor and providing an air storage cavern, a single turbine unit can be used to generate electrical power from 100% of the mechanical output, typically at times of peak power demand. At times of low demand the turbine can be used to recharge the air storage. Worldwide, there are several such systems in operation.At Ragged Chute, North Bay, Ontario, a natural version of CAES using air entrained by a river diverted into a shaft and tunnel system results in a compressed air storage that is used to generate electrical power.Another example is the CAES plant at Huntdorf, Germany.
In Sweden, low level radioactive waste is stored in caverns.
Heating demands in Sweden are high in the winter, despite high building insulation standards. Some district heating schemes are supplemented by water-filled caverns heated by surplus electrical power. This warm water is accessed in the winter via heat pumps.
Sewage treatment plants have been located underground in Scandinavia with obvious environmental and property value benefits.
Book storage for the Royal Library in Stockholm, Sweden is now located in caverns below the original 19th century library building, which had become totally overwhelmed by demands.This elegant building, modelled on the old British Library, has been returned to its intended use with large areas devoted to reading rooms (Figure 5).
Other uses of underground caverns include mass car parking, central telecommunications (civilian and military), bomb shelters, ice rinks and swimming pools.The key benefits are additional security, thermal insulation, reduced insurance costs and maximisation of urban space.
Most appealing of all to the audience at Burlington House was the example of the central store for Aquavit, wine and spirits (Figure 6).
Questions Independent consultant Derek Knight asked about any deleterious effects at the oil/rock interface due to relief of horizontal stress. Lindblom replied that as far as he was aware there were no such problems.There were however some biological problems at static oil/water interfaces due to bacteriological action.The remedy was to minimise water bed areas where possible.
Bob Pine (Camborne School of Mines the author of this report) asked about the impact of caverns in dewatering the soft marine clays common in Stockholm and other parts of Scandinavia.Lindblom said this was well recognised and very strictly controlled in Swedish regulations. For example the allowable inflow to tunnels and caverns was normally limited to a maximum of 1 litre per minute per 100m length.
George Reeves of Newcastle University wanted to know more about the Forsmark store for radioactive waste.Were water curtains part of the containment system? Lindblom replied that the storage of such wastes was indeed different from most of his examples.The host rock must essentially be designed to take care of itself without engineering input. It was not acceptable to rely on a system such as a water curtain for long-term reliability.
Independent consultant John Arthur asked about the relative costs of caverns for LPG storage under pressure or by cryogenics. Lindblom said this depended on the whole gas delivery and storage system. In the most common case, where gas was delivered by ship as a cooled liquid, the cryogenic option was usually the cheaper.
Jan Hellings (Maunsell) asked two questions.With caverns in clay (eg Boom clay LNG example) were there problems with ground settlements/heave or any other problems? And were there any regulatory problems with cavern storage in Scandinavia?
On the first point Lindblom said there could be problems with settlement/heave in clay, but only if the cavern was too shallow.There had been no gas migration problems.On the second question, he said the Scandinavians were now so used to caverns as an obviously beneficial storage alternative that there were few regulatory problems. It could be different elsewhere.A good example was a recent inquiry in Brazil for a frozen LPG cavern.The investigation and design were complete and ready for construction but when the inquiry was asked whether the cavern blasting would trigger landslides, it was sufficiently alarmed to refuse permission to excavate despite the really insignificant risk.
Bob Pine asked how quickly the hot water heat storage system for district heating became effective. Lindblom replied that it usually took two or three annual cycles before the system reached full efficiency but a useful amount was available in the first cycle.
John Arthur asked about the economics of underground caverns compared with alternatives and what were the typical site investigation requirements? Lindblom replied that as long as the cavern did not require substantial systematic support or major treatment to reduce groundwater inflow, it was usually economical. Site investigation was usually in about five stages:
1 Desk study using existing reports and first estimate of the cavern location 2 Feasibility stage design 3 A single hole at the chosen site, including bulk permeability measurements and borehole seismics and any other low cost investigation possibilities.
4 Location of the cavern almost exactly 5 Full design and location adjustment based on a comprehensive site investigation involving about eight holes.
John Harrison of Imperial College asked about differences in practice between UK and Sweden. It appeared to him that underground caverns were the choice of first resort in Sweden and last resort in UK.Was this cultural, or a matter of conditions of contract?
Lindblom replied that in Scandinavia the conditions of contract were less adversarial than appeared to be the case in the UK.Philosophically it was better to have a live design that responded to ground conditions as encountered (on the basis of an adequate but not excessive site investigation) and in which all conceivable problems had not been designed out. By the flexible approach it was possible to have a cavern cost as low as one third of the contractually 'watertight' design.
However, this approach made demands on all parties (client, designer, contractor) to behave reasonably.
Halcrow's Nick Swan asked about monitoring and maintenance requirements for storage caverns. Lindblom said that good monitoring was a big challenge but the current trends in South East Asia (where cavern construction was most active) were encouraging.Operators did want to monitor cavern performance and there was some excellent data management software available.Typically there were not enough monitoring sensors installed, but this should improve.The monitoring should include not only operational but also rock mechanics parameters (deformation, pore pressure, temperature).
With regard to maintenance, most caverns had to be designed knowing that re-entry was virtually impossible because of the dangers of air/hydrocarbon mixtures in the cavern atmosphere.
Consultant Derek Knight asked what was the worldwide ratio of oil to gas storage caverns. Lindblom replied that in terms of numbers of caverns, the ratio was about five oil to one gas.
Ivan Hodgson of Scott Wilson asked about experience with caverns in highly seismic areas, such as Japan. Lindblom replied that experience was very good. Large seismic shocks had been withstood by many caverns, with no apparent damage. The only concern might be shearing of joints/faults if these were highly stressed or seismically active. Most caverns manage to avoid this.
John Dickinson asked about the experience with the Canvey Island cryogenic gas storage didn't this show that frost heave was a problem? Lindblom replied that of course the Canvey Island storage was not a rock cavern and was located in soils at and near the ground surface. It was not surprising that heave was a problem in a shallow high water-content material.Kevin Privett had knowledge of the site and noted that the problem had been exacerbated by using steam on the outside to try to halt the advance of the ice wall!
Barry Clarke (University of Newcastle) asked if there was much potential to use the many existing underground spaces in the UK such as abandoned mines. Lindblom replied that undoubtedly several mine locations were in questionable geology, and, additionally, perhaps in the wrong physical locations. Bob Pine noted that in a recent study in the UK it was concluded that coal measure rocks in general were not suitable for large cavern storages.
John Powell (Arup Energy) asked what were the approximate costs of cavern excavation. Lindblom replied that clearly access tunnels, shafts and top headings were expensive but benching/bulk excavation quite cheap. For a 60,000m 3cavern in Scandinavia the higher costs would be about £80/m 3to £100/m 3and the bulk excavation about one tenth of this.