Clyde Dam ~ The Slip-Joint Problem
The Clyde dam "slip-joint" has been hailed as a design and engineering triumph. When dam workers discovered a faultline during foundation work in 1982, little was known about the geotechnical characteristics of the site, and although instability was a known issue in the Cromwell Gorge, the underlying reasons for this had not been investigated in any detail.
So what exactly did the dam workers discover?
In this case, the fault was essentially a band of pulverised schist rock, running along the bed of the gorge on the right side of the dam (facing). This fine material, between solid rock structures, indicated that substantial movement had occurred on either side.
Some remarkable decisions followed the discovery of the fault, before the full extent of the problem was known. Vast amounts of slurry concrete were pumped into "shear-pin" tunnels drilled across the fault. This grouting was intended to lock the fault blocks together under the dam foundations.
The River Channel Fault, however, is 12-15 kms deep, and critics maintained that the grouting would simply be torn apart, along with the surrounding rock, in a large earthquake. The grouting was described as "dental concrete."
The dam was rather hastily re-designed in 1982, so much so that a sluice channel was omitted from the plans, and the dam workers, of course, built the dam without the missing sluice gate. A later "work around" solution cost $2 million and reduced the dam's generating capacity from 612MW to 432MW.
The most important feature of the re-design was the "slip-joint." It was not based on a known and tested design. However, it was innovative and was regarded as "state-of-the-art" geotechnical engineering. As such, the designers won an award.
So what exactly did the dam workers discover?
In this case, the fault was essentially a band of pulverised schist rock, running along the bed of the gorge on the right side of the dam (facing). This fine material, between solid rock structures, indicated that substantial movement had occurred on either side.
Some remarkable decisions followed the discovery of the fault, before the full extent of the problem was known. Vast amounts of slurry concrete were pumped into "shear-pin" tunnels drilled across the fault. This grouting was intended to lock the fault blocks together under the dam foundations.
The River Channel Fault, however, is 12-15 kms deep, and critics maintained that the grouting would simply be torn apart, along with the surrounding rock, in a large earthquake. The grouting was described as "dental concrete."
The dam was rather hastily re-designed in 1982, so much so that a sluice channel was omitted from the plans, and the dam workers, of course, built the dam without the missing sluice gate. A later "work around" solution cost $2 million and reduced the dam's generating capacity from 612MW to 432MW.
The most important feature of the re-design was the "slip-joint." It was not based on a known and tested design. However, it was innovative and was regarded as "state-of-the-art" geotechnical engineering. As such, the designers won an award.
Location of the Clyde dam join and "slip-joint"The design premise of the "slip-joint" is that the fault will "slip" sideways and perhaps uplift in a large earthquake. To mitigate such movement, the dam was built in two "halves" with a two metre gap between them, sealed on the reservoir side by a vertical concrete "wedge plug" that is held in place by water pressure. The theory is that this "wedge plug" will allow up to 2 metres of lateral and 1 metre of vertical movement, while the water pressure holds it in place against the dam.
Inside the Clyde dam join looking toward the "slip-joint" wedge
Looking up the wedge
Looking down the wedgeAn obvious issue arises if the movement exceeds the design capability. Research has revealed that both small and large movements have occurred on the site, and have been as much as 9 metres laterally. The Alpine Fault has not had a major rupture since 1770, and one is expected within the next 1-20 years in the order of magnitude 8+. Such an event has the potential to result in more than 2 metres of lateral movement through the dam. This shearing action would render the "slip-joint" ineffective, and water under enormous pressure would be forced through the join, perhaps precipitating a catastrophic failure.
But there is another issue that raises a serious question, to say the least. Was the "slip-joint" design premise correct?
The River Channel Fault is described as a "Secondary Fault" because it branches from the main Dunstan-Cairnmuir Fault that dissects the gorge some 3 kms above the dam. Gerald Lensen, one of New Zealand's leading geotechnical scientists involved in active fault research and mitigation planning, maintained that the River Channel Fault was tensional (apart rather than lateral), and therefore the "slip-joint" was NOT designed correctly.
Lensen's analysis becomes clear when the geotechnical structure is examined. The "slip" movement will naturally occur in the main Dunstan-Cairnmuir Fault, and since the secondary River Channel Fault is at right angles to this, it will pull apart as the main fault moves laterally. This opening process had contributed to the formation of the Cromwell Gorge.
The extent of this tensional movement is difficult to estimate, but any such movement must be a serious issue for a "slip-joint" that was not designed to accommodate any significant tensional movement. In a major earthquake, it is possible that the two metre wide “slip-joint” would simply open up and a serious failure could occur as the two 102m high dam "halves" separate. Such an earthquake could also trigger one or more landslides in the gorge, compounding any dam rupture.
Can we have confidence in the "slip-joint?"
It is disturbing to think that the "slip-joint," accordingly, should have been a "tension-joint," or in other words an expansion joint. The folly of building a large concrete dam on an active fault, with an ineffective engineering solution, is both alarming and potentially tragic. Presumably, the dam builders, and the "slip-joint" designers, have faith in their solution. But it is difficult to understand how experts could disagree on such an important issue.
Given the penalty for failure, and the high degree of uncertainty, there was only one safe and rational solution - not to proceed with the dam. But the "think big" politicians and the dam builders were not prepared to swallow such a bitter pill. They decided to accept the risks, on our behalf.
It is precisely because of the complex nature of geotechnical issues that the International Commission on Large Dams (ICOLD) advises against building concrete dams on active faults.
The fact that the Clyde dam was completed in the face of such risks is an indictment against all those responsible. The fate of the dam is, chillingly, in the lap of the Gods. For safety reasons, there is a compelling case for early decommissioning, but the same attitudes that built the dam would certainly dismiss such a call.
Sadly, it will probably take a major rupture of the Alpine Fault, and a failure of the "slip-joint," to draw attention to this potentially deadly issue.
But there is another issue that raises a serious question, to say the least. Was the "slip-joint" design premise correct?
The River Channel Fault is described as a "Secondary Fault" because it branches from the main Dunstan-Cairnmuir Fault that dissects the gorge some 3 kms above the dam. Gerald Lensen, one of New Zealand's leading geotechnical scientists involved in active fault research and mitigation planning, maintained that the River Channel Fault was tensional (apart rather than lateral), and therefore the "slip-joint" was NOT designed correctly.
Lensen's analysis becomes clear when the geotechnical structure is examined. The "slip" movement will naturally occur in the main Dunstan-Cairnmuir Fault, and since the secondary River Channel Fault is at right angles to this, it will pull apart as the main fault moves laterally. This opening process had contributed to the formation of the Cromwell Gorge.
The extent of this tensional movement is difficult to estimate, but any such movement must be a serious issue for a "slip-joint" that was not designed to accommodate any significant tensional movement. In a major earthquake, it is possible that the two metre wide “slip-joint” would simply open up and a serious failure could occur as the two 102m high dam "halves" separate. Such an earthquake could also trigger one or more landslides in the gorge, compounding any dam rupture.
Can we have confidence in the "slip-joint?"
It is disturbing to think that the "slip-joint," accordingly, should have been a "tension-joint," or in other words an expansion joint. The folly of building a large concrete dam on an active fault, with an ineffective engineering solution, is both alarming and potentially tragic. Presumably, the dam builders, and the "slip-joint" designers, have faith in their solution. But it is difficult to understand how experts could disagree on such an important issue.
Given the penalty for failure, and the high degree of uncertainty, there was only one safe and rational solution - not to proceed with the dam. But the "think big" politicians and the dam builders were not prepared to swallow such a bitter pill. They decided to accept the risks, on our behalf.
It is precisely because of the complex nature of geotechnical issues that the International Commission on Large Dams (ICOLD) advises against building concrete dams on active faults.
The fact that the Clyde dam was completed in the face of such risks is an indictment against all those responsible. The fate of the dam is, chillingly, in the lap of the Gods. For safety reasons, there is a compelling case for early decommissioning, but the same attitudes that built the dam would certainly dismiss such a call.
Sadly, it will probably take a major rupture of the Alpine Fault, and a failure of the "slip-joint," to draw attention to this potentially deadly issue.
