Failed Earth Dam

About 100m length of the 300m dam failed at the upper stream as it reached the last 0.3m stage of required 6m height during construction.   WHY?

The Dam was located on a shallow stream with less than 1m depth of water but it was also a former buried deep valley channel of about 25m deep. The soils are mainly marine deposits and the surface soils are extremely sensitive.

The design specified for the staged construction of 1m earth filling and rested for a minimum one month period  to improve the underlying soils before the next stage can continue on. In another words, a lot of standby and monitoring were required. Instrumentation included inclinometers, piezometers, settlement plates and deep gauge meters. The average slope was about 1:7. Another 4m deep excavation was required at the upper stream to increase the storage capacity of the reservoir behind the dam.

The reasons for the failures:

(a)    The Contractor did not construct the earth dam immediately due to low price of the earthworks, less than 15% of the total cost. They did not want to mobilize the earthwork equipments which cost a lot in rentals to them, such as excavators, tractors, compactors and lorries which involved a lot of standby. So they delayed until nine months later and requested for less rest periods and wanted to construct speedily towards the end of the contract period,

(b)   Many of the instruments were not installed properly and reading could not be obtained accurately or doubtful. The worst was that the results were not submitted for supervision team to evaluate immediately, but often two weeks later. In fact from the results obtained from the Contractor later, piezometers showed the sudden increase of pore water pressures and inclinometers showed increased rate of lateral movements before the failure,

(c)    The Supervision team was not experienced enough to handle incompetence yet vocal Contractor on highly technical earthworks and could not do much to enforce the original schedule. There was pressured to complete the work before the coming predicted drought and finally yielded to the Contractor’s demand. The town had been short in supply of water every year. The Contractor claimed that the slope was too gentle and even betted with his balls if the dam failed. This had psychologically influenced the supervision team and Client to doubt whether the design was too conservative,

(d)   There were some localized peaty soils in buried stream, which were not detected during investigation and properly not completely removed and  initiated the failure,

(e)   Continuous movements of heavy equipments such as compactor, excavator and lorries produced vibrations which together with the filled earth might have caused the underlying soft soils, which have an average sensitivity of more than 15, to lose some strengths.

Lessons learned:

(a)    Highly technical method of construction such as soil treatment methods or instrumentation or grouting should be used in caution as many local contractors and even the supervision teams are not skilled or experienced enough to do such works.

The main element of this earthwork design was rest period, allowing time for the underlying soils to gain strength and pore water pressure to dissipate as load was applied. But standby time costs a lot to the Contractor’s equipments and overheads. There was conflict of interests.

(b)   Use conservative design such as piling, heavy geotextiles or excavated deeper to remove the soft and sensitive soils, if there is no good Contractor in the Market, even though the cost may be two to three times more.

(c)    Firm in supervision and provide stern penalty in the Contract Clause at the beginning for failing to start work on vital works and not to be influenced easily by Contractor, whose main objective is to cut cost and earn profits. Also ignore Client or political pressures,

(d)   Site investigation shall be done more closely to detect any shallow localized buried streams/channels, perhaps 5m centre to centre using cheap Mackintosh probes (detect up to 12m depth and auger holes (extract up to 7m depth),

(e)   Instrument installations are often defective or knocked off during construction. Only really competent geotechnical sub-contractors shall be allowed to install and record readings. Alternative instrumentation proposed by Contractor shall not be allowed. Honesty pays an important part in this instrumentation (Many cheatings in installations and readings in local scene). Damaged or non-functioning instruments shall be immediately replaced.

(f)     Pay great attention to highly sensitive soils. There was doubt on the high sensitivity of soils initially whether the tests were properly done, as it was encountered first time in the local scene. The site investigation was executed some time ago and could not verify or re-test. Further investigation fund shall be allocated where doubts arise.