"Chemical Engineering," _J.R. San. Inst._, No. 2, 1910. S. Rideal.

On adsorption phenomena:

1. "Chemistry of Colloids," Dr. W. W. Taylor.

2. "Chemistry of Colloids," V. Poschl.

3. "Chemistry of Colloids," Zsigmondy and Spear.

4. "Chemistry and Physics of Colloids," E. Halschek.

5. "Surface Tension and Surface Energy," Willows and Hatschek.

SECTION VI.--EVAPORATION

The evaporation of the weak gelatine sols (3-9 per cent.) obtained by the processes described in previous sections into sols of such concentration (20-55 per cent.) that they readily set to a stiff gel on cooling, is now an essential feature of gelatine manufacture, and is one of the most important processes.

In the early days of this industry, manufacturers aimed at obtaining a concentrated sol, as this saved time in drying, and so reduced the possibilities of putrefaction. The advent of evaporation has reduced these possibilities to a minimum, and has also enormously reduced the s.p.a.ce required and the capital outlay needed in the drying sheds. It has, in addition, given the practical advantages involved in dealing up to the last minute with a much less viscous liquor. As the liquors extracted are weaker, the extraction is more complete and the decolorization more easily effected.

The earliest attempts at evaporation were not very successful, partly on account of the prolonged "stewing" which ruined the setting power, and partly because of the poor economy of heat. Thus in the open evaporators the sol was maintained at a high temperature for a long period, and this process only proved suitable for low-grade products.

A great stride forward was made by Howard's invention of the Vacuum Pan.

This made it possible to undertake concentration at much lower temperatures, a most important improvement in the case of gelatine and other organic matters easily damaged by heat. The process, however, was still slow, and the sol exposed to heat for a long time, as must be the case when evaporation takes place in bulk. These disadvantages were still fatal to the production of the highest-grade gelatine. There were also the practical difficulties of entrainment ("blowing over"), in which parts of the sol were carried away by the escaping vapour, and also of "incrustation" which so rapidly reduces the heating efficiency and evaporative capacity of the machine. The vacuum pan, however, presented two decided advantages--evaporation at a low temperature, and, as a corollary, the possibility of utilizing exhaust steam to attain this temperature.

Whilst the vacuum pan was a satisfactory machine for many branches of chemical engineering, the problem of evaporation was still unsolved for gelatine liquor because of the "stewing" involved, until the advent of the "film evaporator," which dealt with the liquor not in bulk, but in a continuous stream. In this way the product was only exposed to heat for a comparatively short time. Many evaporators of this type came into being, and rapid improvement was made in the constructional details. The film evaporators retained usually the advantage of evaporation _in vacuo_, so that it was now possible to evaporate gelatine sols by exposure for a short time to a comparatively low temperature. Of this type of evaporator, the Lillie, Yaryan, Schwager, Claa.s.sen, Greiner, Blair Campbell, and the Kestner machines are well-known examples.

A further advance in solving this problem was the application of the principle of multiple-effect evaporation. The vapour driven off during evaporation possesses of course many heat units, and is of very considerable volume. In multiple-effect evaporators this vapour is used to work a similar evaporator, and the evaporated liquor pa.s.ses immediately into what is practically a second machine, and is further evaporated by the heat from the vapour just driven from it. Such an arrangement would be termed a double-effect evaporator. The vapour from the second effect may of course be similarly used to operate a third effect, and the vapour from this to work a fourth effect, and so on.

Thus, we may have triple effect, quadruple effect, etc., even up to octuple effect. The great advantage of multiple-effect evaporation is in the saving of costly steam. Reavell gives the following figures to ill.u.s.trate the economy thus obtained:--

WATER EVAPORATED PER 100 UNITS STEAM.

-----------+-----------+-----------+-------------- Single. | Double. | Triple. | Quadruple.

-----------+-----------+-----------+-------------- 95 | 150 | 220 | 300 -----------+-----------+-----------+--------------

There is naturally a limit beyond which the capital cost of the machine neutralizes the advantage of steam economy, and it is seldom that octuple effects are used. There are probably more triple effects in use than any other machine.

An essential and important part of the modern evaporator is the "condenser," in which the vapour from the last effect is conducted into water (jet condensers) or over cooled surfaces (surface condensers), with a view to producing and maintaining the vacuum.

A lasting vacuum cannot be maintained without an air-pump, as air is often introduced (1) with the steam, having entered the boiler dissolved in the feed water; (2) by leakage from the atmosphere into the condenser and the connected vacuous s.p.a.ces; and (3) in jet condensers, in solution with the circulating condenser water. That from the first two sources may be reduced, but the third is beyond control: hence if high vacua are necessary, surface condensers are to be preferred. Dissolved air is usually 5-20 per cent. of the water volume, and is least for sea-water.

It should be noted that water leaving a surface condenser is in a very air-free state, and therefore particularly suitable for boiler supply.

Apart from the capital cost of a condenser the chief cost of maintaining a vacuum is in pumping the circulating water, of which up to 70 lbs. is usual per lb. of steam condensed.

If W = weight of steam condensed (lbs. per hour); Q = weight of cooling water circulated (lbs. per hour) T{i} = inlet temperature ( F.) of cooling water; T{o} = outlet temperature ( F.) of cooling water; then T{o} = T{i} + 1050(W/Q)

It will be understood that for high vacua, low temperature of cooling water (T{i}) is more important than copious supply (Q/W). It is advantageous, however, to choose a site yielding plenty of cold water, such as a river or ca.n.a.l side. Otherwise it is often necessary to use cooling towers or spray nozzles. The cooling is by evaporation (= 60 to 80 per cent. of W), cold water replacing that evaporated, and yielding water 75 to 80 F. If T{i} = 80 F. and Q/W = 70, a vacuum of 28.34" is possible, but the 0.34" should be allowed for the partial pressure of the air, determined exactly by the air entering and by the displacement of the air-pump.

Another feature of the modern evaporator is the "heater" or "calorifier," by which the liquor to be evaporated is led in a continuous rapid stream through heated tubes immediately prior to its entry into the first effect. It is the aim of the heater to raise the temperature of the liquor to the temperature of evaporation, and so to avoid this being necessary in the first effect. The heater thus further avoids stewing, ensures steady running, and effectively increases the capacity of a machine.

It is noteworthy that superheated steam is not desirable for working an evaporator. The principle of evaporation by steam is not merely that the temperature of the liquor is raised to boiling point; it is that in the condensation of the heating steam its latent heat is yielded to the liquor being evaporated. To evaporate quickly, therefore, the heating steam must condense rapidly. Hence, as superheated steam has a rate of condensation 20-30 times slower than saturated steam, the latter is much to be preferred. A slight superheating, however, may be justifiable where the steam has any distance to travel before use. It is the fact that it is the latent heat of steam which is mainly utilized which gives steam its great practical advantage over hot non-condensable gases.

Steam in condensing yields an enormously greater number of heat units per lb. than hot waste gases. Steam has also the advantage of more constant temperature.

The capacity and efficiency of an evaporator depends upon a good many factors, some of which are worthy of discussion at this point.

The transference of heat and the amount of evaporation are directly proportional to the mean temperature difference between the heating steam and the liquor being evaporated. These temperatures, however, both vary somewhat, the steam losing part of its pressure and temperature as it pa.s.ses along the heating surface; the liquid generally increases in temperature. The mean difference in temperature, moreover, is not the arithmetic mean between the smallest and largest temperature differences, but is given by the following expressions, which yield results not wide apart:--

If [theta]{a} = temperature difference at commencement; [theta]{e} = " " " end; and [theta]{m} = mean temperature difference;

then

[theta]{m} = ([theta]{a} - [theta]{e}) / log([theta]{a} / [theta]{e})

or = ([theta]{a} - [theta]{e}) / [ n(1 - [nth root of]([theta]{e} / [theta]{a})) ]

This mean temperature difference is in practice usually spoken of as the "temperature head" or "heat drop." It will be clear that this temperature head is increased by using steam at higher pressure (temperature), and by evaporating under reduced pressure. Since most liquids have their boiling points reduced about 40 C. by operating _in vacuo_, the advantage of the vacuum is apparent. It should be remembered that the temperature head has not the same value in any part of the scale: it has more value higher up the scale, because the steam is denser and more heat units come in contact with a given area in a given time. It must also be remembered that whilst the pressure gauge is a most useful indicator of steam temperature, it is not necessarily accurate. The pressure in the hot s.p.a.ce is the _sum_ of the pressures of air and steam, and since the temperature (the important condition) of the hot s.p.a.ce depends upon the pressure of the _steam_, and not on the sum of the pressures, the temperature in a steam s.p.a.ce is always rather lower than would be supposed from the pressure indicated by the gauge.

The transference of heat is influenced by the velocity of both the heating fluid and the fluid being heated over the heating surface. The more rapidly each fluid moves, the more rapid is the transference of heat, because a greater number of particles of both fluids are brought to the heating surface in any given time. This is popularly known as the effect of "circulation," and is ill.u.s.trated by the advantage of stirring a liquid being heated in bulk. In the film evaporators the circulation is through tubes at high speed (up to 2 miles a minute), and the maximum effect in this sense is thus obtained. The increase in heat transference is not directly proportional to the increase in velocity, but in a lower ratio, sometimes approximately the square root of the velocity. In such a case, if either velocity be quadrupled, the heat transference is doubled. Other advantages of high velocity are that the heating steam more readily sweeps away condensed steam from the heating surface, and the high-speed film similarly "scours" away "incrustations" on the interior of the tubes.

The transference of heat is also proportional to the conductivity of the metal forming the heating surface. For gelatine liquors, copper tubes are almost invariably employed, the advantage being great even when price is taken into consideration. The following conductivity coefficients ill.u.s.trate this point (calories per hour through 1 sq.

metre of metal 1 metre thick, with a temperature difference of 1 C.):--

Copper...330

Iron.....56

Steel....22-40

Tin......54

Zinc.....105

Lead.....28

The coefficient of heat transmission decreases the more with increasing thickness of wall, the worse conductor is the metal. For copper tubes, however, this decrease is usually unimportant.

The transference of heat is also much influenced by the viscosity of the liquor being evaporated; the greater the viscosity, the lower the coefficient of heat transmission. Unfortunately for this process of evaporation, gelatine sols are exceedingly viscous, and thus the difficulty in obtaining a concentrated sol is thus greatly enhanced.

The transference of heat is often greatly hindered by incrustations of the tubes, which incrustations generally conduct heat very badly. Thus the relative heat conductivities of copper and chalk are as 1000:5.

The amount of heat transferred is of course determined also by the area of the heating surface. The amount of evaporation needed thus determines the number of tubes (of standard size) in the evaporator, and thus the capacity of the machine. An evaporator should have its heating surface area chosen with a view to the duty required of it.