Canal collapse probe tells of small leak

Press, 11 March 1983, Page 26

Canal collapse probe tells of small leak

PA Rotorua Consultants employed by the Wheao canal contractors. Mills Construction, Ltd, have concluded that the December 30 collapse happened after a small leak in ignimbrite rock formed a cavern in the canal wall. In a report presented to the Wheao Committee of Inquiry. Mills Construction told of the findings of two consulting engineers engaged after the collapse to investigate technical aspects of the failure. The two were Mr Robert Morris, consulting engineer, of Christchurch, and Mr M. Mitchell, consulting geotechnical engineer, of Hamilton. The report said that it was considered likely that the event which prompted a leakage on November 23 also prompted the December 30 leak. “As with the November 23 leakage, a small amount of water is thought to have seeped through to the joints within the ignimbrite and then flowed downwards from this point. “Once the initial flow path had been established, erosion in the form of cavern development (from the bottom upwards) would have occured “Eventually the roof of the cavern approached to within a few cemtimetres of the canal water and collapsed occured. “The pumic sand, with its lack of cohesion would have developed an ever-widening hole, in contrast to the tunnel void which has occured in the November 23 event, where only ash soils were present.” The report said that it was likely the cavern collapse prompted a minor slope failure in the canal slope immediately upslope of the cavern. “Gradually soil from this slope failure disappeared into the cavern and through into the voids within the ignimbrite. “Eventually, the top berm of the embankment fell into the void and the embankment was overtopped.” The consultants said that it was necessary to look at the weakest point of the structure to find where the leak started, and this was the ash lining-forebay interface. “There is the possibility that the concrete structure, having warmed by the sun, caused some drying of the adjacent ash. In doing so, some of the water tightness

of the ash may have been lost This could have occured as the fill was being placed in successive layers." "Changes in the canal water level (for example, during dewatering) may have also caused differential movement in the ash lining. The rigidity of the concrete immediately adjacent to the ash would have accentuated any shrinkage movement." The differential movement between the ash and the concrete would have been accentuated at the internal comers of the concrete structure, where compaction of soil was difficult. “However, similar design details occur in most hydraulic structures constructed in other parts of New Zealand with apparently no adverse effects. Possibly, minor leakages do occur at this location in hydraulic structures constructed elsewhere, but there the consequence of such a leakage has not been significant," The consultants said that in the failure mechanism which they had postulated only a minor amount of seepage along the ash-con-crete interface would have been necessary to precipitate the failure. The dominant feature involved being loosely filled joints in the ignimbrite below the ash lining. The consultants said that they had considered closely the competency of the ash lining used elsewhere. along the canal and were of the opinion that further study may be needed before it could be determined whether or not the material was adequate. “In particular, the consequences of minor leakages occuring at various points along the canals must be established.” The provision of butynol or similar lining of the canal within about 100 m of any concrete structure should prevent a recurence of the problem which started the failure. “However, we would suggest that complete lining of the canal with butynol should only be contemplated if the ash itself is proven to be adequate.” This would require further extensive field and laboratory testing, but would be well justified in view of the substantial costs involved with butynol lining the complete length of the canal.