Sawing Slots Into a Growing Dam

The Mactaquac generating station in New Brunswick, Canada, has a problem its engineers cannot fix, only stay ahead of. Its concrete is swelling from alkali-aggregate reaction, and it announced itself in the plainest way a dam can.

In 1985, Spillway Gate 10 was found to be jammed closed as a result of AAR-induced concrete expansion of the intake and spillway east end pier.

Curtis and others, "Practical Evaluation of AAR-Affected Mactaquac Dam," ICOLD, 2021.

That is the signature of an AAR dam. The concrete does not just crack. When it is restrained, as it is everywhere inside a dam, it cannot expand freely, so instead it builds enormous compressive force, and that force squeezes the structure out of shape. Gates bind in their slots. Turbine and generator shafts drift out of alignment. Piers tilt. The dam is, very slowly, tearing at its own machinery from the inside.

You cannot stop the reaction, so you manage the force it makes. At Mactaquac the answer since 1988 has been to cut vertical slots straight through the concrete with a diamond wire saw, relieving the built-up compression and handing the expansion somewhere to go, so the gates seat and the shafts realign. The catch is in the next sentence of the record.

As concrete growth continues, periodic re-cutting of slots is required and most current cuts are made with a 15-mm diamond wire saw.

Curtis and others, ICOLD, 2021.

The slot is not a repair. It is a relief valve that closes. As the concrete keeps growing, the cut faces creep back together, and when a slot has narrowed to a few millimetres it has to be sawn open again. Some slot locations at Mactaquac have been re-cut many times over. In 1995 the work moved into the powerhouse, where transverse slots were cut between the units and, over the following years, the penstocks themselves were cut and re-coupled one at a time to let the building settle back toward true. None of this touches the reaction. It manages what the reaction does.

So AAR management is not a project with an end date. It is an operating discipline. The structure is wrapped in instruments, invar-rod extensometers, plumb-lines, joint meters, that measure the growth continuously, a numerical model turns those readings into a forecast of where the next slot must go and when, and a standing independent review board has overseen the programme for decades. The dam is run the way a chronic illness is managed: monitored, medicated, and never declared cured.

Because it cannot be. AAR needs three things together, reactive silica in the aggregate, alkalis in the cement, and moisture, and in a finished dam the first two are locked in for good. Only moisture can still be influenced, by draining and coating and sealing the concrete, and that slows the reaction. It does not end it, because the heart of a large dam sits permanently against the reservoir and never dries out. This is why the current international guidance, ICOLD Bulletin 184, is titled Management of Expansive Chemical Reactions in Concrete Dams and Hydroelectric Projects. The word is management, not repair, on purpose.

India is not exempt from this. Rihand Dam in Uttar Pradesh, a concrete gravity dam commissioned in 1962, developed AAR within its first decade and has been under repair and watch ever since. And here is the gap worth knowing: India has a good standard for keeping AAR out, IS 2386 Part 7, which screens an aggregate for reactivity before it is ever used, but no dedicated national code for managing a dam that already has the reaction. For that, Indian engineers reach for the same ICOLD guidance the rest of the world uses. With more than 5,000 large dams in the country, about a third of them concrete or masonry and many now decades old, that is a gap the next generation of dam owners will feel.

The lesson: If you inherit an AAR-affected dam, you have inherited a management programme, not a repair job, so set it up like one. Instrument the structure before you think you need to, and measure the growth in real numbers. Define the re-cut trigger by how far the slot has actually closed, not by the calendar. Cut moisture where you can to slow the reaction, knowing you cannot stop it. And budget for a standing programme, with a review board and a forecast model, that keeps the gates seating and the turbines turning for decades. The aim is not to cure the dam. It is to stay one cut ahead of it.

Read more: Alkali-Aggregate Reaction in Dam Concrete →