NZSM Online

Get TurboNote+ desktop sticky notes

Interclue makes your browsing smarter, faster, more informative

SciTech Daily Review

Webcentre Ltd: Web solutions, Smart software, Quality graphics

Feature

Underground Hydro Lakes

Why flood 25 square kilometres of native forest for a hydro scheme when you could put the lake under the forest?

by Earl Bardsley

The greatest environmental impact of a new hydro power scheme is often associated with the water storage component -- new lakes drown existing landscape, natural shorelines are submerged by raised water levels. Yet such impacts may reflect nothing more than a need to create an operational storage volume which takes up only the uppermost few metres of the lake concerned. Could this functional water store be located elsewhere so as not to affect the surface environment in any way?

A simple alternative would be to excavate caverns within nearby mountains and hold the storage water there, giving effective invisibility to the volume fluctuations associated with the daily and seasonal cycles of power demand. Economic realities rule out this option, but the environmental attractiveness of subsurface hydro storage suggests that there are other alternatives worth investigating.

One possibility would be to seek out suitable geological formations, such as aquifers, which can store and transmit reasonable volumes of water through the small pore spaces between the constituent rock particles. A functioning aquifer hydro storage system of this type would alternate between phases of water input and water output --  recharging and discharging -- depending on the balance between power demand and water supply.

How Rocks Hold Water

There are two mechanisms by which geological formations store and release water. The most effective mechanism is water table storage, where an initially "dry" aquifer has its empty pore spaces filled with water. Think of pouring water into a dry goldfish bowl full of marbles -- as the water is added, the water level (water table) in the bowl rises, with the water stored in the "pores" between the marbles. Water can be drained from water table storage by drainage, making the pores empty again.

In real aquifers, these cycles of recharge and discharge might involve transfers of considerable volumes of water. For example, a saturated clean sandstone with a porosity of 0.33 is, by definition, two-thirds rock and one-third (33%) water. This amount of stored water is equivalent to a storage lake one-third the volume of the whole aquifer. In principle, this entire volume of water could be moved in and out of storage at will, although in practice there will always be some amount of aquifer water left at the end of a drainage phase.

The other mechanism of storage is for the aquifer to always remain saturated, but with the constituent rock particles rearranging themselves in response to changes in water pressure as water is forced in or drained out. Less water can be recharged and discharged in this way as the pores are never completely drained at any stage of the cycle.

One possible kind of aquifer hydro storage system could see the water level in a small valley controlled by a local downstream dam. When surplus water is available, it is diverted into the valley and the water level rises. This leads the water pressure to increase in an extensive recharge/discharge network of horizontal tunnels and tributary bores. The raised water pressure in this network causes recharge of the aquifer and forces the water table to rise. The reverse process occurs when extra water is required for power. The water surface in the valley is lowered and the remnant water pressure in the aquifer discharges water back down the bore/tunnel network. Thus the whole water table rises and falls over a wide area.

A variation on the theme would be for the water to be supplied from a pressure tunnel, avoiding the need to flood a local valley. Also, the extensive tunnels in such a network could be replaced by horizontal bores some kilometres in length.

Energy Loss

There is an operational energy loss which will be common to all types of aquifer storage schemes. Work is expended by the gravitational pushing of water into the aquifer, and this energy cannot be fully recovered during discharge phases. The energy loss is evident by the fact that water departs the storage system at a lower elevation than at arrival. This lower head will cause reduced power output per unit volume of water during discharge phases if the output water feeds directly into a penstock. On the other hand, maximum generating head will apply whenever a recharge phase is operating. That is, the water not directed to aquifer storage will bypass to the generators with no loss of head.

The operational energy losses are at least in part offset by the fact that aquifer hydro storage systems can be much more environmentally benign than storage lakes. In fact, by using horizontal subsurface recharge and discharge designs, it is entirely possible that the land surface above would show no visible signs of any alteration. The only ecological effect would be produced from the initial lowering of the natural water table down to the operational elevation of the scheme. This might be of concern for the special case of wetlands, but even wetlands would remain largely unmodified if they were held up by a local impermeable silt layer or iron pan.

The ideal geographical and geological setting for such a system would take the form of an extensive plateau or plain, perhaps bounded by hills to reduce leakage. If the aquifer is being considered as an alternative to a specific storage lake, then the plateau or plain needs to be at about the same elevation as the intended lake to give similar hydraulic behaviour under conditions of gravity recharge.

Our ideal aquifer would be located just beneath the ground surface and be sufficiently permeable to allow the rapid import and export of water via a recharge/discharge network of bores and/or tunnels. At the same time, the aquifer would allow the network to be sparse enough so as to make its construction cost-effective. The storage and release of water would be achieved predominantly by water table changes, so there should be no impermeable layers within the range of water table fluctuation. We might perhaps also add a requirement that the whole aquifer should be resting on impermeable bedrock so as to eliminate vertical leakage loss.

With so many constraints, it might seem unlikely that suitable locations for aquifer hydro storage could ever be found. Interestingly, the necessary fortuitous combination of geology and topography appears to exist on the Blackburn Plateau in the Buller region -- what would be the southern shoreline of the proposed extensive lake associated with the controversial Ngakawau power scheme.

The plateau is slightly higher than the 340-metre level of the proposed Nga-kawau lake. This gives the possibility of avoiding the lake option altogether and instead storing the water beneath the plateau surface. The recharge/discharge mechanism might be something like that considered above, with horizontal subsurface gravity recharge of the sedimentary aquifer which underlies most of the plateau.

Of course, the hydraulic properties of the aquifer and its thickness distribution would have to be studied in some detail to determine the viability of such a scheme. Consideration would also have to be given to long-term sustainability. In particular, it would need to be demonstrated that the system could operate consistently between the extremes of becoming clogged with silt on the one hand and causing subsurface erosion through net sediment export on the other.

From a purely economic viewpoint, an aquifer storage scheme on the Blackburn Plateau is less attractive than the larger storage volumes associated with the proposed lake in the original scheme. The smaller surface area of the plateau would probably mean that seasonal cycles of storage and release could only be realised to a limited extent, although peak daily discharges should be readily achieved. Full utilisation of the Blackburn Plateau might also require the creation of a small lake on the upper Blackburn River in order to reduce leakage. Furthermore, some of the water-transporting rock tunnels in the original scheme would need to be extended to link through to the plateau aquifer.

However, a viable plateau option could produce almost as much power as the original scheme, assuming that the aquifer storage system could operate within a 20-metre range of water table fluctuation. And of course the environmental impact would be negligible compared with the 25-square-kilometres of native forest to be destroyed under the original proposal.

In the end, the decision whether to establish an aquifer storage scheme on the Blackburn Plateau or anywhere else will come down to the economics of construction cost as much as anything. However, given the wide range of possible recharge/discharge configurations it would be surprising if some form of functioning aquifer hydro storage system does not eventually emerge in environmentally sensitive areas in New Zealand and overseas.

Dr Earl Bardsley is senior lecturer in hydrology at Waikato University.