configuration. When the same pumps were used in an even larger unit, a slightly lower total power figure was obtained. The data recorded for the 30,000 GPD unit are representative of conditions at the end of over 2000 operating hours. The data fairly represent the running record with two exceptions. First, during 90% of the operating time the apparent stack resistance was 15% lower than at the end and, second, the current efficiency average over 2000 hr was 85.6 __+ 2%. Allowing for these two factors brings the estimated minimum power to 1.00 kwhr/1000 gal in good agreement with the figure found in operations with the 6000 GPD unit. Conclusions and RecommendationsA self-contained device for the acidification of desalination plant feed streams has been developed. The acid generator requires as inputs electricity and the desalination plant feed stream. Substantial units of 6000 and 30,000 GPD have operated at a total power level of 2.5 kwhr/I000 gal. There are indications that this could be cut in half. The large systems tested suffer from two correctable defects. First, the cells in future units should be long thin rectangles or lozenges to get better flow distribution and to avoid problems associated with buffer precipitate settling in the quiescent portions of the round cell sections used here. And, second, further work on the catholyte/buffer system is required. Flow rates, stream ratios, and the like must be readjusted to produce a high salinity buffer stream before any significant cell power reduction is possible. Further, if the pH of the buffer can be lowered by only 1 unit, additional improvements in power requirements can be realized.

No abstract

No abstract

as expected, fall beloxi the present data T h e measured and calculated data for the investigation are reported in Tables I and 11. T h e standard deviation between k , ellcc, and k , exptl was 0.31 and the average per cent error based upon k, ex,,tl wds 9 4%.Some of this variation can be attributed to error resulting from the calculation of t l T h e results of the study presented as dimensionless J factor correlations suggest that the analogy holds reasonably well for flo\v transverse to vertical tubes in a bundle Nomenclature C = specific heat of mixture, B.t.u./lb. F. do = diameter of tubes, ft. D, = diffusion coefficient, sq. ft. 'hr. G = mass flow rate of mixture, lb.!'hr, sq. ft. min. cross GM = molar flow rate. lb. mole,'hr. sq. ft. min. cross section h, = gas film heat transfer coefficient, B.t.u./hr. sq. ft. F. h,' = total heat transfer coefficient less that provided by gas J o = diffusion J factor J H k = thermal conductivity of mixture, B.t.u.;'hr. sq. ft. k , = mass transfer coefficient! lb. molejhr. sq. ft. a t m . M , = molecular Iveight of vapor, Ib./mole section film, B.t.u./'hr. sq. f t . F. = heat transfer J factor ' F.:'ft. pBJf = mean partial pressure of noncondensable, atm.pt = partial pressure of vapor a t interface, a t m . p,, = partial pressure of vapor in bulk stream, atm. t,= temperature of bulk stream, ' F. t i = temperature of liquid-gas interface, ' F.

No abstract

No abstract

No abstract

The electrochemical oxidation of ethylene to acetaldehyde and carbon dioxide is investigated at a palladium electrode in 2 N sulphuric acid at 8O"C, as a first step towards the partial oxidation of higher olefins. Results on the corrosion of a palladium electrode in the absence of ethylene show features not previously reported, and a possible mechanism is given to account for the results. In the presence of ethylene at electrode potentials above 0.75 V EH, it is shown that two reaction schemes are likely to be in operation to' account for the products observed, as well as the appearance of a considerable amount of palladium metal in the electrolyte. / Preparative runs were carried out in the larger cell. In this case the anode was folded into a W shape to get a large surface area. The anolyte was stirred by gas bubble agitation.Both cells had platinum counter electrodes. 912889