Symposium on leak management 30 March 2004

Presentations

Prevention is better: the design of watertight grout layers

In his lecture, Van Tol discussed the design of a watertight grout layer. It became clear that a guaranteed watertight design is not possible in practice, or is unaffordable. A certain optimum between permissible risk and cost will always have to be chosen. This was investigated by TU Delft by applying a failure probability model for a grout layer, analysing issues such as misalignment of the injection lance and location determination as possible sources of error.

In practice, the hydraulic resistance of a layer is always smaller than the resistance you would theoretically expect based on the material properties. Small defects have a big impact on the 'over all' resistance of the system. Conclusion: if you achieve 50% of the design resistance, then you have delivered a good product. However, it is then difficult to determine whether or not that 50% will lead to a calamity. If the flow comes evenly through the grout layer, then the risk is zero. If it comes through a single hole, then there is no problem as long as there remains enough cover and the hole remains filled with soil material. Only when excavation is carried out to just above the leak is a calamity, as the sand is forced out of the hole and then the flow rate (with material transport) increases explosively. Van Tol advocates installing 'dormant drainage' to deal with such calamities.

Questions from the audience mainly concerned these restrictive measures, such as applying dormant drainage. These dewaterings should be short in duration, to prevent damage to the surrounding area, but allow the builder to carry out remedial measures. In general, the competent authority will cooperate.

Leak management at the Griftpark in Utrecht

Leurink does not like to talk about Leak Management, because it sounds so negative. But practically speaking, the cement-bentonite wall built around Griftpark in the 1990s is indeed leaky.

But: in environmental circles, only zero emissions are permissible, and they are met. So, how so: leaky? Also, everything was realised within budget, so: the Griftpark is a success, even with leakage.

Leurink outlines the history of the site, the (polluting) activities that have taken place there and the geological soil structure, which shows that the Kedichem clay, to which the cement-bentonite wall connects, is not present throughout the site. A prognosis (with uncertainty limits) was prepared in advance for the flow to be extracted and treated. This distinguished between vertical (through the Kedichem layer) and horizontal (through the wall) inflowing water, with the contractor to be charged for the latter component. In practice, this did not happen because the overall system performed within the predetermined uncertainty interval. As a result, some more water was pumped and purified than calculated beforehand, but it fell within the uncertainty margins and did not result in adjustments to the abstraction system, purification or pipeline capacity.

The discussion revealed that the requirements set in the tendering procedure may have led to an overly expensive design.

Implementation aspects

According to Admiraal, the starting point must be that leaks cannot always be prevented, but that the risks must be managed.

First of all, steel sheet pile walls are discussed: causes of leaks can include obstacles, 'heavy' foundations that require heavy vibratory hammers and sometimes the need for reinforcement. The contractor can be prepared for these aspects through soil investigation and historical research, and Admiraal also mentions a number of things the contractor can do to reduce the risk of leakage in practice. These range from working carefully, via lock indicators and pre-drilling or fluidising to applying all kinds of fillers to make the locks less permeable.

Admiraal then similarly discusses a number of other techniques to apply water-reducing walls, such as grout piles and diaphragm walls.

After vertical walls, horizontal layers such as:

1. natural inhibitory layers
2. underwater concrete
3. artificial jet grout layer
4. foil structures

Admiral's conclusions are:
Leaks occur easily but are difficult to control, therefore one must be aware of the risks and determine beforehand whether leaks are acceptable. In extreme cases, alternative construction methods that pose less risk should be adopted. He urges open communication between all parties involved.

Discussion: again, following this lecture, the need to weigh up the risks of possible leakage at an early stage and have precautionary measures in reserve is again raised. Nevertheless, it remains a difficult balancing act. The consequence is that investments (e.g. dormant drainage) are made that ultimately do not have to serve.

Leak detection: locating the problem

In his lecture, Slob discusses geoelectric techniques. These techniques are the most appropriate when it comes to leak detection, especially as they are reasonably fast. In particular, he elaborates on the ECR/EFT method recently introduced on the Dutch market by Texplor. In a way that is clear to everyone, Slob shows how the techniques in question work and what the possibilities and limitations are as a result. It is especially fascinating to learn that things like dimensions of leaks and depths can potentially be determined with this method. This does require additional effort. He also introduced listeners to a simpler version, in which measurements are taken while walking through the field, as it were, instead of the fixed set-up, which is rather labour-intensive.

The responses show that the methods in question do not have widespread familiarity, even less is known about how the method works. There is a rather subjective approach regarding the confidence placed in the method.

Leak repair

Firstly, the question is raised as to whether there should always be recovery. This was also raised in earlier lectures. The answer is "No", only when there are risks. While one such risk is water encumbrance, Lambert advocates not being too rigid about it. Water heaviness should be an aspect in the consideration, not a sacred number. Two other risks are settlement and hence damage, a slow process; and piping and hence soil transport, a disastrous and rapid process that should certainly not be solved by extra lowering of the water level. This only accelerates the process.

Options to prevent settlement damage include return pumping, when predetermined groundwater levels in the surrounding area are undercut. Sometimes this is a relatively simple and inexpensive measure.

In addition to traditional repair methods, which (when applied at some depth) are without exception expensive and labour-intensive, and also require accessibility to the exact leak site, there is a need for quick, simple repair methods that, while not leading to complete closure, reduce the water load to acceptable proportions.

Lambert describes a method under development: 'Biosealing', using the properties of subsoil, groundwater and leak. Feeding the micro-organisms naturally present in the soil causes them to multiply rapidly, leading to clogging. Formation of iron sulphide makes the blockage more durable. These mechanisms have been tested in the laboratory; a field experiment is currently underway.

The discussion looks at the durability (FeS formation) and the explanation of the mechanism. There is also a cautionary note: to what extent is it to be feared that tarnishing of the (steel) sheet piling will occur.