The following text may have been generated by Optical Character Recognition, with varying degrees of accuracy. Reader beware!

48 Fig. 11, illustrates the plant cov- ered by the patent of C. S. Lenz and G. E. Jupp, in which the water flows downward through a succession of annular channels into bulky spaces where the oil disengages itself and re- turns upward through collecting chambers suitably spread. The re- moval of the oil may, however, be con- siderably impeded by the flow of water in these oil-collecting spaces. Fig. 12, shows the “H. and H.” de- hydrator, in which the collecting plates analogous with those in Lenz and Jupp’s apparatus are caused to rotate by the impinging jet of the mixture supply. This gives a slight accentuation of gravity, tending to throw the water to the outer an- nular clearance spaces, the value of which as a separating agent has to be offset against the mixing effect of the distributor. In this apparatus a float automatically controls the oil discharge, as in the “Pirbright” sepa- rator, already described. The latest form of separator (in point of date) is that produced by the Stream-Line Filter Co. Fig. 13. This is of the horizontal flow type, with a closed system and a quiet oil-collecting chamber from which oil is withdrawn from time to time on observation of the inspection cocks. After passing through a preliminary chamber, in which the bulk of the oil is separated, the main flow traverses a large number of small tubes in parallel, of such proportions that steady flow is induced. In these tubes the small gloubles of oil are enabled to separate. In one type they pass through holes in the tops of the tubes, and, from the quiescent water around, into the oil-collecting chamber; in another type they are allowed to collect in the tubes until they coagulate and are carried forward by the water to a trap where—being now in a readily separable form— they easily emerge from the water flow and pass into the ‘oil-collecting chamber. Eddy Motion In Oil Separators ‘In this discussion of particular separator designs nothing has so far been said of the means for avoiding eddying motion, although it has al- ready been mentioned that in an ed- dying flow the smaller oil globules can never be separated by gravita- tional means. Referring now to Figs. 1 to 13, it would appear that in many separators the baffling arrangements are on the whole liable to promote rather than reduce eddying, although even this is of secondary importance when it is realized that for plants of the MARINE REVIEW size and capacity used in practice even the flow through a smooth cir- cular pipe (the best possible ar- rangement for steady flow) would be of an eddying nature. The cri- terion for steady streamline flow in a circular pipe is well known, be- vd ing that * (Reynolds’s function) B/ p must not exceed a critical value, which is, in point of fact, about 2000. Sub- stituting the value of 4/p) for water at normal temperature gives v d -|- 1/40 ft. ?/sec. *a = velocity, d = diameter, »& = viscosity, p = density. Now the volumetric flow Q through Td < V, and a pipe of diameter d is 4 substituting from above this gives Q 7 a sf wd | — xX — ft. ?/sec.—i.e. —— ft. */sec., 4 40 160 if d is in feet. Thus the permissible throughput if eddying is to be avoided increases di- rectly as the diameter (and not as the square of the diameter, as might have been anticipated), and if a throughput of 200 tons per hour is required, d must be given by 200 tons 35:4.” hour wd x x =— hour ton 3600 sec. 160 ft.*/sec. 200 x 85 & 160 8 = «100 FE. aout 3600 7 If eddying is to be avoided in a single circular pipe a throughput of 200 tons per hour necessitates a diameter of 100 feet and for any other form of cross-section the dimensions would need to be still further in- creased. It is obvious, therefore, that in any separator not providing for a highly divided flow, eddying must occur with reasonably large throughputs unless the dimensions are made unreason- ably large. Further, since it is much easier to keep below the critical value of vd when d is small, small-scale models of any reasonable form of separator give satisfactory results, so that subsequent failure on a large scale is at first a matter for sur- prise. It was in order to overcome the difficulties associated with eddying mo- tion that the writers proposed the type of apparatus (now manufac- tured by the Stream-Line Filter Co. Ltd.,) shown in Fig. 18, in which the large number of small tubes en- sures that the flow in each is of a steady streamline character. Reference must now be made to the September, 1926 various processes which have been proposed with the object of removing the last traces of oil from the water before discharge. All known forms depend for their action on the adhesive nature of the oil, the filtering material being such as offers little resistance to the flow of water, but retains the oil either on its surface or in its pores. There is a certain amount of secrecy as to materials actually em- ployed, but it is well known that some makers have used felts of cow-hair or wool, while others have tried wool flock, cotton-wool etc. These may be classified as of organic origin; while, on the other hand, an inor- ganic material is used by the Stream- Line Filter Co. In developing a suitable filter two main difficulties have to be met: 1. The large quantities of water to be dealt with. oily 2. The necessity for renewing the filtering material economically. With regard to 1, an _ efficient preliminary separator will limit the quantity of oil to be dealt with by the filter to less than one part in ten thousand of water, though even this means twenty gallons of oil per thousand tons of water. It is, ‘there- fore, clearly very desirable to elimi- nate as much oil as possible in the separator, either by making it very large or by the means suggested by the writers. With regard to 2, the replacement of filtering material is costly, so that means for cleaning are practically essential. Organic fibers can be cleaned by a dry-cleaning process, al- though this involves removal from the filter and is not an operation which would commend itself to shipowners. Inorganic material is readily cleaned by blowing back with steam; the in- crease of temperature reduces. the adhesiveness of the oil and it is therefore removed by the eurrent of steam. This method is inapplicable to organic materials because of the resulting shrinkage and _ structural changes. Summary of Present Position As stated already, it is possible to reduce the oil content in the final water discharge to not more than one part in two hundred thousand of water with a plant of reasonable size, which fulfills the conditions of being easy to operate, self-contained, and obviating expensive renewals. With such a plant the prime cost can be written off in a few months by the value of the oil recovered, and the (Continued on Page 56)