US2141049A - Gas holder water treatment - Google Patents

Description

Dec. 20, 1938. J. R. SKEEN 2,141,049

GAS HOLDER WATER TREATMENT Filed Oct. 14,v 1937 2 Sheets-Sheet l v'r T A If M. h i 1J' Lax/m cw @1123 INVENTOR ATTORN EY Dec. 20, 1938. R, SKEEN 2,141,049

GAS HOLDER WATER TREATMENT Filed Oct. 14, 1937 2 SheetsSheet 2 INVENTOR ATTORNEY Patented Dec. 20, 1938 GAS HOLDER. WATER TREATMENT John R. Skeen, Philadelphia, Pa., assignor to The United Gas Improvement Company, a corporation of Pennsylvania Application October 14, 1937, Serial No. 168,911

14 Claims.

This invention pertains generally to the control of bacterial growth and pertains particularly to the reduction or prevention of discoloration of surfaces and corrosion of iron in iron-water systems.

The invention will be described in connection with gas holders used in the gas industry. However, it is to be understood that it may have many other applications.

For many years the gas industry has been concerned with the appearance of gas holders and has sought to make such holders as inoffensive in appearance as possible.

As a result the'exposed surfaces are painted, partly for protection, but largely for appearance.

This presents no particular problem in the case of pressure and waterless holders since the maintenance of the paint is no more difficult than the maintenance of the paint on any other structural surface.

In the case of water sealed holders, however, an entirely different problem arises. Surfaces which are intermittently immersed in water with the rise and fall of the lift or lifts acquire a reddish brown discoloration, very unsightly in appearance. In warm weather it may require only a few weeks for a freely painted surface to be almost completely masked with a reddish brown film.

The industry has heretofore merely resorted to palliatives such as matching the paint to the discoloration, spreading oil on the water surface, mechanically skimming the water surface, or illtering the water without arriving at the cause of the trouble.

The rusting of iron in water or moist air is of universal occurrence. If the water in contact with iron is stagnant, the surface of the water becomes covered with a heavy, tenacious rustyappearing scum. Although circumstances make a water sealed gas holder a particularly suitable place to study the phenomenon, it occurs in the same way in a bucket of water or in any other similar environment.

To illustrate the problem as applied to gas holders, reference will be made to the drawings in which:

Figure l is a sectional elevation diagrammatically illustrating a single lift gas holder; and

Figure 2 is a sectional elevation diagrammatieally illustrating a multiple lift gas holder.

Referring now to Figure l, a simple form of water sealed gas holder is shown as consisting of an inverted bell l0 extending downward into a tank H containing water l2 so that the inverted u holder is used.

bell III is free to rise and fall according to the amount of gas confined in it.

A guide frame for the bell II) as it ascends and descends is illustrated at l3, and pipes for the introduction and withdrawal of gas above the level of the water l2 are illustrated at M and I5 respectively.

To increase the capacity of the holder without increasing the size of the tank, the multiple lift This is illustrated in Figure 2, wherein a bell 20 and lifts 2!, 22 and 23 are shown in an expansible-collapsible arrangement which extends down into tank 26. A guide frame is illustrated at 25 and pipes to introduce and to draw off gas above the level of the water 26 are illustrated at 21 and 28 respectively.

The bell 20 and lifts 2i and 22 have attached to their lower peripheral, edges outwardly extending annular cups 29, 3|] and 3| respectively, and lifts 2|, 22 and 23 have attached to their upper peripheral edges inwardly and downwardly extending grips 32, 33 and 34 respectively.

As illustrated in Figure 2, bell 20 and lift 2| are full of gas. A seal is formed between bell 20 and lift 2| as the result of the dipping of .grip 32 into water 35 in cup 29.

If more gas is now forced into the holder, the lift 2| will be raised entirely out of the water. In rising, cup 30 containing water 36 dipped from tank 24 rises underneath grip 33 with grip 33 entering the cup 30 to form a seal with the water 36 therein.

As more and more gas is forced into the holder each succeeding lift 22 and 23, and as many others as may be provided, rises in turn until the holder istfull.

'As gas is withdrawn from the holder the lifts and the bell descend in the inverse order in which they rose. 40

It will be noted that the water level in tank 24 is such that the cups and grips engage and disengage below the water level.

The bell, lifts and framework are usually constructed of steel. While steel tanks are in general use and are usual in newer constructions, tanks have also been constructed of concrete and of brick.

In cold climates it is necessary during the winter to heat the waterin the tank and cups to prevent it from freezing. This is usually accom-- plished by introducing steam, or less often hot water, into the cups and into the top of the tank.

The holders so far described are conventional in character.

It will be obvious that a gas holder may have any desired capacity ranging from the small gasometers used in laboratory work up to the large holders used for supplying metropolitan centers. Holders up to two hundred and fifty feet in height have been built with tanks up to three hundred feet in diameter. The capacity may be up to twenty million cubic feet of gas or more. The water contained in the tank may be as high as twenty million gallons or more. i I

It is not customary to paint the insides of the bell and the lifts and consequently large surfaces of unpainted iron come into contact with an enormous volume of water and particularly that within the boundaries of the lower lift.

Since one or more lifts are always at least partly down except during the short intervals when and if the holder is completely filled, and

the lower lift is completely down during a large part of the time, for the purposes of convenience in description the waterwithin the submerged lifts, that is, the water directly underneath the gas, will be referred to as the inside water and the water in the annular space between the submerged lifts and the tank, that is, the'water exposed to the atmosphere, will be referred to as the outside water".

Incidentally it should be noted that the latter is relatively small in volume compared to the former, isof the same depth, and has a small annular surface area exposed to the atmosphere compared to the large circular surface area of the inside water which is exposed to the gas.

The inside water, beingin equilibrium at its surface with the gas, is for the most part almost devoid of oxygen, even though oxygen is substantially soluble in water, because of the 'very small oxygen content of the gas. On the other hand the inside water carries a relatively high carbon dioxide content because of the relatively high carbon dioxide content of the gas and the substantial solubility of carbon dioxide in water.

Contrasted to this since the outside water is in equilibrium at its surface with air it carries a relatively high oxygen content and arelativelylow carbon dioxide content. This is, of course, becauseair has a relatively high oxygen content and a relatively low carbon dioxide content. I

At intermediate points between the surface of the inside water and the surface of the outside water, the oxygen content and the carbon dioxide content will be intermediate the values at the two surfaces.

This is, of course, assuming that there are no disturbing conditions such as rainfall, convec-,

tion or other currents, etc. However, on the whole and for the purposes of this description such conditions may be assumed to exist at least for practicable purposes.

Although iron is relatively insoluble in water,

actually minute quantities of iron go into solution in water. Due to the absence of oxygen and the presence of carbon dioifide in the inside water this iron takes the form of a ferrous carbpnate namely Fe (H003): or Fe'COa, according to the following equations,

tration of ferrous carbonates is reduced in some way.

As stated above, currents of various kinds cause a certain degree of mixing of the inside and outside water. It will be appreciated, for instance, that the raising and lowering of the lifts produces considerable churning calculated to cause mixing. This is particularly true of the lower lift which upon being raised also reduces the boundary between the inside water and the outside water.

Furthermore, the laws of diffusion apply to this system, tending to bring the inside water and the outside water to the same "composition.

\ As a result of the foregoing, inside water containing ferrous carbonates in solution becomes mixed in the annular space outside the lowered lifts with the outside water which carries oxygen in solution, and one or more of the following reactions take place.

. It will be noted that ferric hydroxide is a product of each of these reactions.

Ferric hydroxide is characterized by its insolubility in water, by its yellowish brown color, and by its flocculent or colloidal nature.

While'ferric hydroxide may be dehydrated to the oxide according to the following equation 2Fe(OH)3=FezO:+3Hz0, this is not thought to take place to a great extent due to the large volume of water present which would tend to drive the latter reaction to the left. At most, hydrated oxides F82O3(H2O)n are formed. However, all of these compounds'are very insoluble,

are deeply colored and may take part in the distion of the inside water is not appreciably effected,

first, because the total volume of inside water is so many'times greater than that of the outside water, and second, because a large part of such oxygen is lost to the vapor space above the water in establishing equilibrium conditions.

Such outside water, however, serves to unsaturate the inside water in ferrous carbonates by dilution and consequently more iron is dissolved.

If the highly colored ferric hydroxide, ferric oxide and/or hydrated ferric oxide would settle out of the outside water at a rapid rate, there would be no problem of' holder discoloration.

Unfortunately the cycle is complicated by the presence of a micro-organism belonging to the plant family commonly designated as iron bacteria. ,The group has the peculiar habit of thriving in the presence of iron.

Green plants obtain their energy for growth from sunshine and almost all other living things obtain their energy for growth from green plants.

One of the few possible exceptions is the group iron bacteria, the members of which may obtain their energy for growth from the oxidation of ferrous iron to ferric iron, although it appears that they cannot thrive unless oxygen is also present.

Although iron bacteria are well known, it was not known prior to my invention that they were concerned either with holder discoloration or, as far as I am aware, with the ordinary processes of iron rusting.

I have reason to believe that the particular micro-organism involved, which I prefer to call ugiensis, differs from the forms of iron bacteria previously identified.

That an organism associated with the well known phenomenon of iron rusting has apparently remained unidentified for so long a time can be explained by a combination of facts. First,

face of the outside water where oxygen is in solution in relatively large quantities, its growth decreasing with depth because of decreasing concentration of oxygen. Colonies adhere to unpainted metal surfaces suchas the inside surfaces of the tank and of the submerged lifts (but perhaps not of the bell) below the water line.

The corrosion of iron is greatly accelerated: by the absorption of ferrous carbonates from solution and their deposition within the plant body as hydrated ferric oxides, and due to stimulation of iron dissolution from those surfaces immediately beneath the colonies by the acidic reaction resulting from their normal metabolism. I v v The solution ferrous iron collects in the gelatinous sheaths around the cells of the organism and when oxidized to the ferric form produces an iron colloid with the dead cell materials.

It is this iron-organic complex that gives rise to the continuous gelatinous scums on the surface of the outside water.

As a lift emerges from the water during the inflation of the holder the layer of scum sticks to the paint. I

This scum is extremely difiicult to remove while still wet and upon becoming dry cannot be removed without destroying the paint film.

As a matter of fact the scum may be likened to a paint in which the iron rust constitutes the pigment, the organic debris the vehicle, and the water the thinner.

Each time a lift of the holder is immersed and then raised this skimming process is repeated until the paint underneath is eventually completely masked.

A secondary phenomenon of this organism also contributes to the disfiguration of the holder paint: The peripheral cells of a colony adhering to an unpainted metal surface become brittle with accretions of ferric iron. Agitation causes dislodgment and a yellow sediment thus formed diffuses through the outside water-toward the liftsasit settles to the bottom of the tank, and some of it becomes lodged on rivet heads and other projections on the lifts.

This sediment, upon drying, adheres almost as tenaciously as the dried scum.

If it were not for the iron bacteria the insoluble ferric iron compounds would settle to the bottom of the tank without causing any serious discoloration.

Having ascertained the basic cause of the dimculty, I have solved it by creating an unsuitable environment for the growth of the organism or organisms responsible for .the discoloration.

I find that increasing the pH of the water, for instance by the addition of sodium bicarbonate, reduces but does not prevent scum formation. Furthermore, even such weakly alkaline solutions attack many paint films so severely as to desroy them quickly.

I find further that the addition of common bactericidal reagents such as phenol, arsenic, copper, mercury salts, etc., along with an increase in the pH of the water does not affect an improvement over the above since the variety of iron bacteria involved is found to be extraordinarily resistant to the common chemicals that are toxic to most forms of life. Various types of the newer fungicides and bactericides proved equally ineffective.

Conditions most favorable to these or anic growths are warmth and a neutral or slightly alkaline water. The optimum temperature appears to be about 85 F. which explains why more trouble has been experienced during the summer months.

The propagation rate decreases rapidly as the solution becomes more alkaline or more acid.

For instance, should the water be maintained with a pH of 4.5 or less, the growth of the organsm would not be troublesome, but unfortunately the rate of iron corrosion is greatly increased.

Similarly should the water be maintained with a pI-l of 8.0 or more, the rate of growth would be reduced (but not suppressed), but unfortunately this is not only severe on paint films but is also somewhat diflicult and expensive to maintain since the absorption by the water of carbon dioxide tends to prevent the maintenance of a high pH.

Upon investigating the growth of the organism in solutions of varying pI-l, I have discovered that only very small concentrations of chromate ion obtained for instance by the addition of sodium dchromate are required in solutions of above '7 and say between 7.4 and 7.9 pH, to control satisfactorily the growth of the organism even at the optimum temperature of 85 F.

I find that a very satisfactory way to maintain a desired pH in the presence of chromate ion is by the addition of the required amount of a suitable substance such as sodium bicarbonate or sodium carbonate.

I find further that within limits the concentration of either component may be reduced if the amount of the other component present is increased. In this case I prefer that the product of the two added salt concentrations exceed a given minimum value and find that this will keep organic growthbelow the scum forming level.

Curiously enough the chromate ion reaches its opt-mum effectiveness at a pH which in the absence of 'chromate ion is distinctly favorable to growth of the organism.

On the other hand if sodium dichromate were used alone it would not only be ineffective but would practically disappear'from the water in a few weeks.

In combination the two ions (chromate and hydroxyl) show an effectiveness which is unequalled by any other known treatment.

When the dichromate is first added to water '(which usually contains at least some oxidizable components in solution or suspension) there is vary these figures which are given merely for theonly a small loss due to reduction after which very little further loss is experienced.

Since treatment of such large volumes of water requires comparatively large quantities of reagents, cost becomes a consideration.

From this standpoint I wish to point 'out that a small amount of growth is permissible without causing noticeable discoloration and that the water need not be so toxic to the organism as to prohibit completely its growth.

Furthermore, the presence of some chromate ion under such circumstances that the pH is greater than 7, and particularly when greater than approximately 7.4, will have a measure of beneficial action in that it will reduce to some extent at least the rate of growth of the organism even though it may not be sufllcient to prevent the formation of scum altogether.

Furthermore, I find that with increase in holder size smaller and smaller concentrations of the reagents are su'fllcient to prevent discoloration. I believe this to be due to a decreasing ratio of total area of unpainted surface to the total volume of water present. In other words, as holders increase in diameter, the area of the inside surfaces of the lifts and unpainted surfacw of the tank do not increase in proportion to the volume of water in the tank. Moreover, the width ofthe annulus between the tank wall and the lifts increases with increase in size of the holder.

As an example, for instance, in the case .of large holders, that is holders of a gas capacity considerably greater than two million cubic feet,

I find that growth on unpainted metal surfaces may be permissible up to about one per cent of the area of such surfaces on which growth takes place, whereas in the case of holders of a gas capacity of considerably less than two million cubic feet such growth-should possibly be restricted to about 0.15 per cent to obtain absolute assurance against any possible discoloration.

It is understood, of course, that conditions may centraticns-the relative amounts of each can be varied over a rather wide range.

This is illustrated in the following table in which minimum recommended concentrations are given, not only individually but also in terms of their product.

line show ranges from the smallestto the largest .holders of the particular holder size bracket, and are based on a permissible iron corrosion of 16% of normal for the smallest holder to 20% of normal for the largest holder of the bracket;

Similarly the values for the third horizontal line were obtained, the permissible iron corrosion being less than 16% of normal.

It will be noted that the product of the minimum values given for sodium bicarbonate and sodium dichromate does not equal the recomprinciples for obtaining complete protection against discoloration with the minimum amount of chemical additions.

I prefer to initially add chromate ion in the I form of sodium dichromate. suitablechromium compoundmay be employed.

As chromate ion slowly disappears (by reduction and overflow)- its. concentration, may be maintained, if desired, by the addition of chromic acid, by the addition of more sodium dichroinate,.

or by the addition of any other suitable substance. Chromic acid might be used in the first instance, except that at present it is considerably more expensive. v

I prefer to maintain the desired pH of the water by the addition ofsodium bicarbonate, or a sodium bicarbonate yielding substance such as sodium carbonate. I find for instance that cost may be reduced by adding sodium carbonate, which is less expensive.

The relatively high concentration of ,carbon dioxide in the inside water, which as above explained is maintained by the presence of carbon dioxide in the confined gas, will gradually convert the sodium carbonate into sodium bicarbonate.

However, since sodium carbonate will produce a higher pH which will be more severe on ordinary paints than the pH produced by sodium bi- TABLE Batlio water R800 d dmini R ded ni vo ume mmen e eeommen ml mum Gas capacitiy (If holder in galloillstlo1 fi g gsi ggg mfg mum lgglllfigiligot godgufitgi gTHalgJ O; and;

on o co un colleen on percen a; i 2 pereen sm soes in in percent by weight by weight concentrations square feet 2,000,000 d up 300 320 x 10- f 24 x 10- 2.28 x 10- 2,000,000 ownto300,000 311M025 320xl0- to400x10' 24X10 t032X10' 2.28Xl0' t03fllX10'P Below 300,000 I 25 400 x l0- 32 1 10- 320 x 10.

In explanation of the table, the first column lists ranges of holder sizes which of necessity are arbitrarily chosen. Accordingly, the values given in thelfirst horizontal line for holders having a gas capacity of 2,000,000 cubic feet and up are the minimum values for a holder of 2,000,000 cubic feet gas capacity. These values are based on a permissible corrosion which is 20% of normal, and may be decreased with increase in holder size following the principles above set forth.

Likewise, the figures in the second horizontal carbonate, with gases such as oil gas, natural gas or other gases containing much lower concentrations of carbon dioxide than coal gas, producer gas or water gas, it may be necessary to use bicarbonate initially, unless, of course, a special alkali resisting paint is used on the surfaces which are submerged and also painted.

Any other suitable substance or substances may be substituted for maintaining the desired pH of the holder water.

Attention is, however, directed to the fact that However, any other withdrawal.

many substances which might be otherwise suitable for the addition of chromate ionand for the maintenance of the desired pH, have components present which might tend to function as plant oods.

For instance, it is commonly known that certain phosphorus, potassium, nitrogen, calcium, iron and magnesium compounds are plant foods.

Therefore, in selecting a compound to be substituted consideration should be given to its possible effect upon the stimulation of growth of the organisms which it is desired to control orv destroy.

In selecting such substitute compounds consideration must obviouslyalso be given to the question of solubility. and compatability as well as relative cost.

In connection with the question of compatability, consideration should be given to the fact that there are present in the gases substances such as cyanogen and hydrogen sulfide and that reactions with these substances is usually to be avoided in the holder.

When the final selection is made the substances chosen may be added to the holder water in any desired manner.

I have added carbonate and dichromateby separate solution in water followed by pumping into the holder tank.

Some consideration should be given to mixing with the holder water.

I have accomplished this by drawing the mixing water in batches from the bottom of the tank and when almost saturated with the substance being admitted pumping it back into the tank at a point within the lifts, near the top, and near the opposite side of the tank from the point of The process is 'repeateduntilthe prescribed mass of treating substanceshas been added.

Any other method of adding the treating substances may obviously be substituted.

Thorough mixing, however, isioi' importance to the life of the chromate. v

If sodium carbonate is'used instead of the bicarbonate and if the paint on the lifts is not alkali resisting, I recommend that care be taken as to the length of time that the lifts are immersed in order to avoid destruction of the paint. The reduction in pI-I due to conversion of the sodium carbonate into sodium bicarbonate by carbon dioxide as well as the efiiciency of mixing should be checked at suitable intervals until at least a substantial part of the normal carbonate has been converted into bicarbonate.

Since practically all holder waters contain calcium and magnesium salts which are precipitated .by sodium carbonate, I recommend that the lifts be fully inflated during the addition of the treating substances to prevent discoloration by deposition of the precipitated materials on painted surfaces.

I have found that an average of about two weeks is required for complete clarification of the water from such precipitated substances.

The treated water will enter the cups upon immersion of the lifts and I recommend that attention be given to the dilution of the water in the cups as a result of rainfall or of steaming to prevent freezing, in winter. Circulation of tank water through the-- cups might be provided, or the lifts might be immersed, or the water in the cups might be brought back to the desired composition by any other suitable means.

It appears that the chromate will react with lead paints unfavorably. Therefore, I recomcovered with two or more coats of lead-free paint in which case no trouble will usually be encountered.

There are now available many lead-free paints which may be used for primers and top coats.

During the initiation of the treatment the organism in a desperate effort to survive develops and grows with amazing rapidity in a greatly modified form and appears as isolated colonies adhering to the unpainted metal. These colonies die, become brittle and disintegrate, causing some turbidity and sediment. The metal underneath the colonies appears etched but fortunately the etching process does not continue for more than a few weeks after the treatment is installed, with no appreciable harm to the metal.

After the treatment has been installed and sodium carbonate has been converted to bicarbonate, it is thought that the inside water near its surface is actually mildly acidic with a pH in the neighborhood of about 6.7. This is because of the high carbon dioxide content of the water. In this case, chromium is probably present as dichromate.

On the other hand, the outside water near,

and for a considerable distancebelow, its surface is distinctly alkaline with a pH determined almost 1 The absence of yellow, and the presence of a clear blue color, in the outside water substantiates this viewpoint.

-In this connection I wish to point out that the organism is capable of growing only in the presence of oxygen, that is, in theoutside water for a limited distance below its surface, say, about fifteen feet. Over this distance the water is not only alkaline but is of the clearest blue, indicating the presence of the chromate ion instead of the dichromate ion.

At the surface of the outside water with the concentrations suggested for a holder of 2,000,000 cubic feet capacity the pH is in the neighborhood of approximately 7.85.

For the purpose of this specification and the claims, the term iron where used is intended to include not only iron itself but the various forms of iron and steel, including alloys susceptible to corrosion by the process of rusting.

While particular reference has been made to gas holders, it is to be understood that the invention is not limited thereto and may be applied generally to the reduction of discoloration and/or corrosion where similar problems are involved. By the term annulus as used in the claims is meant the space in the tank in which the holder water is exposed to the atmosphere at its uppersurface.

. It is, therefore, to be understood. that the particular description. is by way ofillustration and that changes, omissions, additions, substitutions and/or modifications may be made within the scope of the claims without-departing from the spirit of the invention which is intended to be limited only as required by the prior art.

I claim:

1. A method for the reduction of the growth of iron bacteria and the formation of scum in an iron-water-oxygen system in the annulus of a water sealed gas holder containing gas, comprising maintaining said water in an alkaline state in the presence of chromate ion.

2. 'A method for the reduction of the growth of iron bacteria and the formation of scum in an iron-water-oxygen system in the annulus of a water sealed gas holder containing gas and having painted surfaces exposed to said system, comprising maintaining said water with a pH in the neighborhood of 7.4 to 7.9, and with a concentration of chromate ion suiiicient to be materially toxic to iron bacteria.

3. A method for reducing the growth of iron bacteria and the formation of scum in an ironwater system exposed to the atmosphere in the annulus of a water sealed gas holder having painted surfaces exposed to said system, comprising adding to said water substances capable of maintaining the pH in at least the upper part of said annulus above -7 but below 8 when said gas holder contains gas, and of producing chromate ion in solution.

4. A method for reducing the growth of iron bacteria and the formation of scum in the annulus of a water sealed gas holder having painted lifts, comprising adding to the water in the tank of said gas holder sodium bicarbonate and sodium dichromate.

5. A method for reducing the growth of iron bacteria and the formation of scu'm in the annulus of a water sealed gas holder containing gas including carbon dioxide and having painted lifts, comprising adding to the water in the tank of said gas holder sodium carbonate and-sodium dichromate, and maintaining the lifts in raised position until said carbon dioxide in said gas converts at least a substantial portion of-said sodium carbonate to sodium bicarbonate.

6. A method for reducing growth of iron bacteria and the formation of scum in the annulus of a watersealed gas holder having painted lifts, comprising adding to the water in the tank of said gas holder sodium bicarbonate and chromic acid.

7. A method for reducing growth of iron bacteria and the formation of scum in the annulus of a water sealed 'gas holder containing gas including carbon dioxide and having painted lifts,

comprising adding to the water in the tank of said gas holder sodium carbonate and chromic acid, and maintaining the lifts in raised position until said carbon dioxide'in said gas converts at least a substantial portion of said sodium carbonate to sodium bicarbonate.

8. A method for reducing discoloration of the paint on the lifts of a water sealed gas holder, comprising maintaining the water in the tank of the holder alkaline where exposed to the atmosphere, and adding to Said water one of a group consisting of chromate ion and dichromate ion.

9. A method for reducing discoloration of the paint on the lifts of a water sealed gas holder comprising maintaining that portion of the water in the tank of the holder between the inside wall of the tank and the outside walls of the submerged lifts and sufliciently near the surface to absorb any substantial quantity of oxygen with a pH above 7.0 and below 8.0 and with sufficient chromate ion in solution to be substantially toxic to iron bacteria.

10. A method for reducing discoloration of the paint on the lifts of a water sealed gas holder comprising maintaining that portion of the water in the tank of the holder between the inside wall of the tank and the outside walls of the submerged lifts and sufliciently near the surface to absorb substantial amounts of oxygen with 'a pH sufficiently high to convert dichromate ion to chromate ion but not greatly in excess of 8.0, and adding dichromate ion to the tank water.

11. A method for protecting the paint on the submerged lifts of a water sealed gas holder from discoloration comprising maintaining the water in the annulus between the wall of the tank and the lifts and sufilciently near the surface to absorb substantial quantities of oxygen with a pH below 8 but sufficiently high to convert dichromate ion to chromate ion, and adding sufflcient dichromate ion to the tank water such that when converted to chromate ion in the above mentioned portion of the annulus conditions are sufficiently toxic to reduce the area of growth of iron bacteria on unpainted iron surfaces to at least 1% of the total area of unpainted iron surfaces upon which such bacteria are capable .of growing.

12. A method for reducing discoloration of paint on the lifts of water sealed gas holders of 2,000,000 cu. ft. capacity and below comprising adding to the water of the tank hydroxyl ion at least equivalent to the addition of 320 x 10- by weight of NaHCO3, adding to said water one of a group consisting of chromate ion and di chromate ion at least equivalent to the addition of 24x 10-% by weight of Na2Cr-2O'z-2H2O, and increasing at least one of such additions so that the product of the 'equivalentconcentrations of NaHCOa and NazCizOq-2Hz0 will be at least 2.28 x 10- 13. A method for reducing discoloration of paint on the lifts of water sealed gas holders of gas holding'capacity of at least 2,000,000 cu. ft. comprising adding to the water of the tank hydroxyl ion approximately equivalent to the addition of 320x10 by weight of NaHCOs, adding to said water one of a group consisting of chromate ion and dichromate ion approximately equivalent to the addition of 24 x 10 by weight of NazCrzOv-ZHzO, and increasing at least one of said additions so that the product of the equivalent concentrations of NaHCOa and NazCrzOv-2Hz0 will be at least approximately 2.28 x 10- 14. A method for reducing discoloration of the outer surfaces of. the lifts of water sealed gas holders, comprising maintaining that portion of the water in the annulus which is sufllcien y near the surface to absorb any substantial quantity of oxygen with a pH at least approximately between 7.4 and 7.9 and with sumcient chromate ion in solution'to reduce the area of growth of iron bacteria on unpainted iron surfaces to at least 1% of the total area of unpainted iron surfaces upon which such bacteria are capable of growing.

JOHN R. SKEEN.

Images