【正文】 nd between lifts at that point. Fuel spills were a problem at Monksville Dam because the contractor was allowed to refuel on the dam. That potential problem was addressed at Elk Creek Dam by specifying that the contractor use a movable refueling pad （墊）on the surface of the dam to collect and contain fuel spillage. The contractor chose instead to refuel equipment and perform maintenance off the surface of the dam. Mixing RCCIn most cases, the methods used for transporting, spreading, and pacting RCC will not affect production as much as the speed and efficiency of mixing. Thus, the mixing plant capacity should exceed the laydown capacity(設計容量).Achieving the desired product consistency and quality at continuous, high production rates requires good concrete plant design and rigorous maintenance. RCC mixes are relatively harsh and sticky, and the lean, dry material has no fluid properties. The design of misers, transfers, and hoppers must take these characteristics into consideration in order to avoid caking（結塊）and loss of capacity.The mixing method must produce a homogeneous mixture of the ingredients. This factor determines the mixing time and, to a large extent, the production rate. Tests should be done to determine retention times（拌合時間）required for each mix. The variablespeed pugmill mixers （葉片式攪拌機）used during 1987 at Elk Creek Dam required 38s of mixing for 6yd3 () batches. At Willow Creek Dam, the retention time in the 9yd3 ( m3) drum misers was 75 s.The trend in design is to use one RCC mix for the body of the dam. For large projects that require a variety of mixes, the plant chosen for the job must be able to change mix designs quickly and little or no mechanical or manual manipulation to plant ponents. On most projects, different plants are used to produce conventional concrete and RCC.For major projects the weights of all ingredients should be digitally recorded as a function of time, date, and mix design. Because of the sensitivity of RCC to excess water, the plant should be equipped with instrumentation to determine the fine aggregate moisture content. In Japan, automatic waterbatching systems are adjusted according to results of continuous monitoring of the moisture content of sand in the sand storage piles. This System has been used on most Japanese RCD projects and has resulted in precise control of moistllre colltent. In the United States most contractors feel continuous visual inspection of the RCC placement or a Vebe test is a more reliable control of water content than moisture meters in the aggregate piles.Proper blending or ribboning of the aggregates and cementitious material on the charging belt as they are fed into the miser will help to speed mixing time and avoid buildup of the sticky material. Achieving the proper timing and angle of introduction of water into batehed mixtures is also important. The proper sandwiching of material to achieve best results with RCC often is different from conventional mass concrete mixes. Each plant and mix bination has its own quirks（特點）, so the exact method of ribboning mix constituents（要素） can only be determined by trial and error(反復試驗).Highspeed batching and mixing can best be achieved with individual weigh systems for aggregate. Accumulative weigh systems make it more difficult to finetune the ribboning of constituents on the belt. Mixing plantsVarious binations of batch and continuous concrete plants using drum and pugmill mixers have been used to produce RCC. Continuousmix plants can provide higher output capability than batch plants, and the most sophisticated continuous mixers can produce the same degree of control as batch plants. Land use and labor requirements are generally less for continuous plants.Drum mixers have the advantage of using less energy than pugmills and provide good control of batch proportions. Pugmills are faster and are generally more portable than drum mixers. Pugmills have been used on all RCC dams in Japan. Pugmills need to be carefully designed to avoid maintenance problems and excessive wear on paddles and plates, particularly for mixes with large aggregates. In drum mixes, excessive buildup of the sticky material can occur, reducing capacity and mixing efficiency. Redesign of fins（鰭狀物） has helped to avoid this problem in some cases. Considerable cleanup of the drums on a daily basis is often required.For highvolume production, largerthanaveragecapacity drum mixers may be necessary to cope with problems encountered when mixing noslump RCC. Blending a zeroslump (塌落度)mixture with a high proportion of fines often causes long mix and discharge times. At Willow Creek, a fourbin Noble 600 plant with two 9yd3() Erie Strayer drums was used. The plant had been proven on an earlier highway job at 600 to 750 yd3(460 to 570m3) times were 30 s longer than for the fluid highway paving mix, and the drums had to be derated（降級為）to between 7 l/2 and 8 yd3 ( and 6 m3) because the diameter of the discharge end was increased to speed dumping. On average, the plant produced 400 yd3/h (310 m3/h) of RCC and peaked during one shift at 438 yd3/h (335 m3/h).Four 250yd3/h (190 m3/h) IHIHydam pugmills made by Ishikawajima Construction Machinery were used by the Japanese contractor for mixing batched RCC at Elk Creek Dam. The doubleshaft(雙軸) pugmills were hydraulically driven and allowed for variable rotation speeds during the charging(進料), mixing, and discharging processes .The coarse aggregate was crushed basalt sized from 3/4 to 3 in (19 to 75mm) maximum. Batching was puterized using ErieStrayer controls. The maximum output of the four 6yd3 () mixers was 1014yd3/h (775 m3/h). That was achieved on October 29, 1988.Two different batchtype mixing plants were used to produce RCC at Upper Stillwater. The first plant built at the site was a Noble 600 with two 8yd3 (6m3) tilting drum mixers. A second plant, a Noble 600 modified to accept two 4yd3(3 m3) Nikko highintensity,