The Art of Dewatering Design

Dewatering and Excavation

There can be a fairly high level of risk in designing water removal systems for excavations. Although that risk will always be present, there are ways to reduce it and allow for the inevitable changes that will occur most of the time.

How to Reduce Risk When Designing Water Removal Systems for Excavations

Nearly all bid contracts require that the design to remove or control the groundwater for nearly all excavations below the water table be part of the contractor’s bid. The design team usually does not provide design details on where the contractor places the wells, their size, and depth. Instead, the contractor must provide a design of the means and methods and meet certain technical criteria that are written into the specifications. Although I have seen responsibility for dewatering system design shift significantly several times in my 35-year career, the problem remains the same. Creating a design and then applying a factor of safety do not seem to work in dewatering.

Several decades ago, the civil/geotechnical engineer or hydrogeologist would provide a design for both temporary and permanent means of lowering the water table. These designs were included as part of the design documents, which the contractor would subsequently bid on and build. If the subsurface conditions changed during construction, the designer would make the necessary modifications to the design and the contractor would build the dewatering system that way. The contractor would typically bid the job as a lump sum for the engineer’s design, with a unit price to add more wells, sumps or interception trenches as needed during construction. This reduced the amount of claims because the contractor had a significantly reduced risk.

However, this approach changed during the late 1980s and early 1990s. By this point, geotechnical engineers and hydrogeologists thought that they understood the subsurface quite well and computer models were everywhere. Thus, they began to treat the design of dewatering systems as an “exact” science based on their modeling, and didn’t allow a way to modify the system by including unit prices for potential changes. This had a cascading effect. When the given design did not function as intended, the result was an increase in claims. Disagreements ensued among the design engineers and the contractors in terms of who was responsible for the errors. The project owners believed that it was the engineer’s responsibility, so the management at A/E firms decided to move responsibility for means and methods to the contractor’s side of the contract.

Now, why is there so much risk in dewatering or groundwater control design, when in most other facets of geotechnical engineering design there is not? It appears that the engineering community has found a way to minimize risk by applying a reasonable factor of safety or overdesign that everyone accepts. But this does not work in the subdiscipline of dewatering because it appears that there are more variables and therefore chances for uncertainty in design or modelling, than in any other. Permeability, for example, is given in orders of magnitude, which means if our prediction is slightly off, the result can double the number of wells predicted.

Over many years of working in the engineering industry, I have discovered several ways to reduce the risk to all and control costs. Put simply, it is critical to minimize variables as much as possible. Here are some ways to do so:

  • Reuse the same well installers/drillers and their methods of building/installing a well. By doing so, the designers know what product they are getting. This will minimize the variability that can occur as a result of different people installing the wells.
  • Design for the inevitability of change. No matter how much experience you have, conditions will not be what you anticipated—for example, the silty sand is cleaner in a particular local area and the well spacing will not work as designed because the drawdown curves do not intersect.
  • If the soil column to be dewatered is layered with different soil types, design for the higher permeability one everywhere. The system will be slightly overdesigned, but if the layers change thickness or permeability, the design will work.
  • Use a system that has the wells closer together so that you are more assured that the drawdown curves will overlap even if the subsurface conditions change.

Where possible, work with a system that allows changes without great cost. I prefer a vacuum dewatering system because, for a limited cost, the system can be modified to achieve the desired results, especially if you “jet,” not drill, the vacuum wells into place. It is worth noting that vacuum dewatering systems are not what they used to be. Good pumps can produce 28 in. of mercury (Hg) vacuum and also pump nearly 2,000 gpm.

This jetting approach has several significant advantages:

  • The jetting approach is a fifth of the cost of drilling.
  • Jet water washes the fines out of the aquifer and makes a developed well quickly.
  • The jet water “head” up the hole keeps the hole temporarily open as the well is quickly installed and the filter is sand packed around it from the surface.
  • The reusable header or collection pipe that extends the vacuum from the pump has many ports, so if a wet area is discovered, an additional well or two can usually be quickly installed by hand.


There can be a fairly high level of risk in designing water removal systems for excavations. Although that risk will always be present, there are ways to reduce it and allow for the inevitable changes that will occur most of the time. Not all designers and/or contractors are the same, so do not treat this specialty as an interchangeable commodity.

Frank   Pita,   PE,   LHG,   D.  GE, is a Principal and Vice President with Jacobs Associates. He has more than 30 years of experience with geologic and heavy civil engineering projects and is an expert in soil mechanics, dewatering and heavy civil shoring.

Comments are closed here.