RSS FEEDS
   Print this page




Mass Concrete and the Benicia-Martinez Bridge
Ric Maggenti and Bob Brignano, California Department of Transportation
Crossing the east end of the Carquinez Strait at the confluence of the Sacramento and San Joaquin Rivers just prior to entering the San Francisco Bay, stands a new 1.4-mile (2.2-km) long Benicia-Martinez Bridge on the lifeline Route 680 in Solano and Contra Costa counties. The bridge has 335 cast-in-place, lightweight concrete, single cell, box girder segments with spans up to 660 ft (200 m) between 11 piers, 10 of them rising out of the waterway. The specified compressive strength for the lightweight concrete was 6500 psi (45 MPa) at 28 days. The bridge is built to withstand any maximum credible earthquake generated from major faults running through the region. The bridge is 82 ft (25 m) wide, accommodating five lanes of traffic and is engineered for future light rail.

Over 100 piles with diameters of 8.2 to 9.1 ft (2.5 to 2.8 m), the massive pier footings, pier walls and columns, and pier tables and diaphragms were cast-in-place normal weight, high performance concrete (HPC). Most of these HPC elements are greater than 6.6 ft (2 m) thick and were treated as mass concrete with thermal control measures being necessary. However, the high-strength lightweight HPC with its lower mass but much higher cementitious materials content resulted in much thinner elements needing thermal control.

The mass concrete temperature control measures were both passive and active. The main passive control measure consisted of lowering the initial concrete temperatures prior to placement although fly ash and coarse grind cement were also used. The former was achieved with the use of chilled batch water, ice replacement of batch water, and liquid nitrogen injection as necessary. Active control was achieved by casting polyvinyl chloride pipes in the concrete elements. During the setting and hardening of the concrete, cold water pumped from the strait was circulated through the network of piping to remove the heat generated by hydration of the cement.

There were over 200 mass concrete placements with normal weight HPC. The measured temperature exceeded the specified maximum of 160°F (71°C) on only two placements. All but a few placements used 3/4-in. (19-mm) diameter cooling pipes spaced at 2 to 5 ft (0.6 to 1.5 m) apart. The normal weight concrete cementitious materials content ranged from 615 to 800 lb/cu yd (365 to 475 kg/cu m). The highest recorded temperature was 165°F (74°C) in a 9.2-ft (2.8-m) diameter pile with a concrete having a cementitious materials content of 792 lb/cu yd (470 kg/cu m) and no cooling pipes.

The 335 single-cell box girder segments, nine hinge segments with large diaphragms, nine midspan closures, a two-span section cast on falsework, and various secondary concrete placements were all cast with lightweight HPC. With a cementitious materials content of 980 lb/cu yd (581 kg/cu m) coupled with a low fly ash percentage of 5 percent and metakaolin at 10 percent, the mix generated more heat than any of the normal weight HPC. Thermal control measures were implemented to limit peak temperatures to 160°F (71°C).

For the lightweight HPC, pre-cooling with ice and liquid nitrogen was necessary for most of the concrete and cooling pipes were necessary for many elements. Thin elements that are not normally considered mass concrete can still reach an undesirable peak temperature if enough heat is generated and it cannot dissipate fast enough to the nearby surfaces. Cooling pipes were used in the thin elements cast with lightweight HPC, with the cooling pipe spacing ranging from 6 to 18 in. (0.15 to 0.46 m)—much less than for the normal weight concrete elements.

Besides the heat generated from the high cementitious materials content, the temperature rise was also higher. The red and blue curves in Fig. 1 show the different peak temperatures of lightweight concrete and normal weight concrete, respectively for 3 ft (1 m) test blocks. The concrete blocks differ only in the type of coarse aggregate. Both blocks have the same cementitious materials content generating the same heat rates in the same environment, but the lesser mass block rises to a higher temperature with mass being the only difference between the blocks. For comparison, the green curve shows the behavior of a large, low-heat generating element cast with normal weight HPC.

A graph shows three plots of concrete temperature in both °C and °F versus elapsed time in days. The first plot is for a 1-m lightweight concrete block and shows a peak temperature of 89°C (192°F) at 0.72 days. The second plot is for a 1-m normal weight concrete block and shows a peak temperature of 75°C (167°F) at 0.80 days. The third plot is for a normal weight concrete footing and shows a peak temperature of 65°C (149°F) at 3 days.

Fig. 1. Comparison of concrete temperatures.


Though only segment elements 3.3-ft (1-m) thick or more were initially thought to need thermal control, after the first few placements and thermal analysis it was concluded that all lightweight concrete elements needed thermal control. This included the 1.8-ft (0.55-m) thick stems and deck elements as thin as 0.92 ft (0.28 m). Overall for the job, thermal control went well. Most of the elements where the temperature exceeded the 160°F (71°C) limit occurred during the first placements as the properties and thermal control procedures for this high strength lightweight concrete mix design were being developed. In fact, 15 of the first 20 segments cast had elements with measured temperatures greater than 160°F (71°C) with four of these exceeding 176°F (80°C).

The highest temperature recorded was 196°F (91°C) in a lightweight concrete segment soffit where only the passive method was used. However after casting these first 20 segments, the measured temperature rarely exceeded 160°F (71°C), while most elements were kept below 131°F (55°C). The frequency curve of peak temperatures of stems, soffits, and decks is shown in Figure 2. Note the higher temperatures of the deck elements although these are the thinnest sections. This is because cooling pipes were not used as often in the deck elements, and most of the peak temperatures were recorded in areas without active thermal control measures. In contrast, only the first 17 of the 335 stem pairs did not have cooling pipes and cooling pipes were not used in about half of the soffits, with those being at the thinnest locations.

A graph shows the frequency of occurrences of maximum temperatures from 26 to 95°C grouped into 5°C ranges. Separate plots are provided for the decks, stems, and soffits. The graph illustrates that the deck, which had thicknesses ranging from 11 to 19 in., had higher peak temperatures than either the 22-in. thick stems or the10 to 59-in. thick soffits.

Fig. 2. Frequency of peak temperatures.


With many factors influencing the characteristics and measures to cope with the heat of mass concrete, ACI Committee 207—Mass Concrete was set up in 1930 to gather information on theory and practice regarding construction of large concrete dams. Since then, the theory and practice of mass concrete has come to apply to much smaller concrete elements made with HPC. Experiences on the Benicia-Martinez have demonstrated the importance of measuring the heat characteristics of the concrete prior to casting the actual structure and using demonstration placements to verify the thermal analysis and proposed concrete temperature control procedures.

HPC Bridge Views, Issue 47, Jan/Feb 2008