What Breaks Concrete From the Inside

Reinforced concrete carries a quiet paradox: the steel inside it does not rust, even though steel rusts almost everywhere else. The reason is chemistry. Concrete's pore water is strongly alkaline, close to pH thirteen, and that alkalinity grows a microscopically thin passive film on the steel that all but stops corrosion. So a corrosion problem in reinforced concrete is really the story of losing that protection. Two things take it away. Chloride, from seawater, brackish water or de-icing salt, works its way through the cover and, past a threshold, breaks the passive film in spots and starts pitting the bar. Or carbonation: the air's carbon dioxide slowly reacts with the concrete and pulls its pH down toward nine, at which the film can no longer hold, and the steel corrodes along its length. Either way, once the film is gone, the bar rusts, and rust takes up far more room than the steel it came from, so it pushes the cover off from the inside.

The Fifth Avenue Dam in Olympia, Washington, shows how this plays out on a structure nobody thinks of as a corrosion risk. It is a concrete flood-control and tide dam, built around 1950, with reinforced-concrete pier walls and wing walls and radial steel gates, and it sits where the lake meets a saltwater inlet, so its concrete lives in brackish, chloride-laden water. When engineers assessed its condition, the surface already told part of the story.

The inspection revealed minor to moderate cracking and spalling in the concrete wing walls and pier walls, with localized areas of advanced spalling and exposed steel near the bottom slab.

Moffatt and Nichol, "Capitol Lake Dam Condition Assessment and Life Expectancy."

But cracks on the surface do not prove the cause. To know whether chloride had reached the steel, they cut cores and measured the chloride at the depth of the reinforcement. In two of ten cores, it was over the threshold where the bar is no longer safe.

When a chloride percent of 0.400% or greater is measured at the level of the reinforcement in a concrete sample, it is reasonable to assume that corrosion of the inner reinforcing steel has been initiated.

Moffatt and Nichol, "Capitol Lake Dam Condition Assessment and Life Expectancy."

That is what turns a visual survey into a diagnosis. The spalling and exposed steel showed where corrosion had already broken out; the chloride profiles showed how far the salt had advanced toward the steel that still looked sound. And the profiles drew a line down the middle of the dam: the cores over the limit came from the Budd Inlet side that faces the salt water, while the cores from the freshwater Capitol Lake side were almost chloride-free, at 0.02 and 0.03 percent. Same concrete, same age; the difference was which water each face stood in. The report rated the pier and wing walls as fair, not failed, and recommended the right kind of intervention for chloride that is already in the concrete: corrosion inhibitors and cathodic protection, which stop the reaction rather than just patching the damage, because you cannot wash the chloride back out.

The lesson: On a dam, the corrosion risk is rarely the mass concrete, which is thick, lightly reinforced and often permanently wet or dry. It is the reinforced-concrete parts in the wet-and-dry, salt-and-spray zone: the gate piers, the trunnion anchorages, the spillway bridges, the intake towers. Protect them the cheap way, before the fact, with ample cover and dense, low-permeability concrete that keeps chloride and carbonation away from the steel for decades. And inspect them for it, because corrosion is silent until the cover cracks. A half-cell potential survey reads the electrical sign of a rusting bar through sound-looking concrete, and a few chloride cores tell you how close the salt has come. By the time you can see the rust, the cover is already coming off.

Read more: Climate Change and the Durability of Dam Concrete in India →