Hot, Warm, Cold, Super Cold

The monsoon stops a pour. The roller-compacted concrete that went down before the rain started is still in place, and nobody on site knows whether the next layer is four hours away or four days. The question that decides what happens next is not whether a joint has formed between the two layers. A joint has formed. The question is what kind of joint it is, and the answer is not the resident engineer's judgment. It is the product of two numbers.

The first number is the temperature of the exposed surface. The second is the time elapsed since that surface was placed. The maturity factor of the joint is the product of the two, with one refinement the Teesta engineers applied: because concrete effectively stops gaining strength below minus 12 degrees Celsius, the temperature is measured from that origin, so 12 is added to the surface temperature in degrees Celsius before it is multiplied by the exposure time in hours. The paper calls this the modified maturity factor. A surface at 28 degrees held for eight hours carries a modified maturity factor of 320. The same surface left for two days carries close to 1,920. The higher the number, the further the lower layer has set before the upper layer arrives, and the weaker the bond between them will be unless the surface is treated.

At Teesta Low Dam Stage-IV, the 160 megawatt roller-compacted concrete dam that the National Hydroelectric Power Corporation built on the Teesta river in West Bengal in 300 millimetre lifts, the engineering team set the treatment classification down by maturity factor. A factor up to 400 was a hot joint. From 400 to 600, an early warm joint; from 600 to 800, a late warm joint. From 800 to 1,600, a cold joint. Above 1,600, a super cold joint. Read as the conditions the paper names in its text, that is four steps, hot, warm, cold, and super cold, with the warm band split in two for treatment timing. Each step up the scale carried a heavier treatment requirement.

The reason the classification matters is documented in the Indian Standard that defines the term by name. IS 7861 Part 1, the code for hot weather concreting, defines a cold joint as "a joint or discontinuity formed when a concrete surface hardens before the next batch is placed against it, characterized by poor bond unless necessary treatment is given to the joint." The phrase that carries the weight is the last one. Poor bond, unless treated. The joint is not automatically a defect. It becomes one only if it is treated as though it were still hot when it is not.

The NHPC team was candid about why this lesson had to be learned the hard way. In the same paper that set out the classification, they noted that three dams had been built with roller-compacted concrete technology in India before Teesta Low Dam Stage-IV, and that the experience had not been clean.

In the early RCC dams, problems posed with seepage through lift joints and shrinkage cracks were reported.

S.K. Yadav, B.N.S. Naveen Kumar, and B.K. Chaudhary (National Hydroelectric Power Corporation), construction paper on Teesta Low Dam Stage-IV RCC gravity dam, CBIP / INCOLD symposium proceedings, 2021.

Seepage through lift joints is the dominant failure mode in a roller-compacted concrete dam, because the horizontal joint between two layers is the plane of least resistance to water moving through the structure. A hot joint, placed and compacted before the lower layer has set, knits into a near-monolithic mass and needs no treatment at all. A cold or super cold joint has to be rebuilt as a bond surface: the laitance and weak surface mortar removed by green cutting as the lower concrete approaches final set, the surface then cleaned and brought to a saturated surface dry condition before bedding mortar and the next layer go down. Where green cutting is too late because the surface is already days old, ACI 207.1R, the American Concrete Institute guide to mass concrete, specifies cleaning by wet sandblasting or by a high-pressure water jet of at least 6,000 pounds per square inch, roughly 41 megapascals.

The Teesta Low Dam Stage-IV team did not solve the joint problem by treating joints better. They solved it by arranging the work so that most joints never went cold in the first place. The dam was split into vertical blocks at contraction joints spaced 25 metres apart, the corners were placed with grout-enriched vibratable concrete, and the pour was kept moving with set retarders so that successive layers arrived while the surface below was still in the hot range. Their summary of the result is the cleanest statement of the principle in the literature.

Keeping HOT joint condition between layer was assured leading to not having to treat the RCC surface before starting the next fresh layer.

Same paper, NHPC, CBIP / INCOLD 2021.

Conventional mass concrete runs the logic in the opposite direction and reaches the same destination. A traditional concrete gravity dam cannot place its next lift for days, because the heat of hydration in the lift below has to dissipate first. IS 14591, the Indian Standard on temperature control of mass concrete in dams, puts the usual minimum elapsed time between successive lifts at 72 hours. That joint is cold by definition, and it is planned to be cold: treated as a designed construction joint, prepared as a bond surface, accepted as part of the structure. The roller-compacted concrete dam races to avoid the cold joint. The conventional dam schedules it deliberately. Both work, because in both cases the joint is classified honestly and treated by its class.

The lesson: When the monsoon stops a pour, the joint that forms is not waiting for an opinion. It is waiting for two measurements. Record the surface temperature and the elapsed time, compute the maturity factor, classify the joint, and treat it by its class. A hot joint needs nothing. A cold or super cold joint needs green cutting, surface preparation, and bedding mortar before the next layer, and treating it as though it were hot is how lift-joint seepage gets built into a dam meant to carry water for a hundred years. That decision is made in the first hours after the rain starts, at the lift, by the engineer with the thermometer and the clock. It is not made at the lab three weeks later. Plan the monsoon, log the interruption, classify the joint, treat it by its number.

Read more: Cold Joint Prevention in Mass Concrete Dam Construction →