Determining Corrosive Effect of Water Supplies
Great Damage Caused by Corrosion—Effective Methods of Testing Water— Use of Spray from Nozzles for Removing Carbonic Acid
THE following paper makes some important suggestions as to the neutralization of corrosion arising from qualities in the water supply and in a practical way treats of the methods of determining the relative corrosibility of these supplies.
One of the most troublesome problems connected with water supply is its injurious effect on metals. While the water may be perfect from a sanitary standpoint, its damage to water and service pipes, boilers and laundry operations, may cause incalculable loss.
In those districts where the water is high in alkalinity and hard, the worst effects are the formation of scale in boilers and the large amount of soap required in washing. A well-known remedy is, of course, to course, soften the water by lime or soda. But where the water is low in alkalinity and soft, or contains free carbonic acid or sulphuric acid, or too much oxygen, the result is corrosion. Generally, this corrosion action is too small to be immediately noticeable, but eventually, the iron pipes tuberculate, and hot water receptacles are subject to pitting.
There are many water works systems, that have supposedly excellent water supplies, which, in a few years, insidiously reduce the capacity of the mains by tuberculation.
Great Damage Through Corrosion
Take the eastern states and all those districts supplied with soft water from the mountains, where cast iron pipe is used, tuberculation is a universal accompaniment, which apparently cannot be overcome by any pipe coating yet devised. In the coal mining regions, corrosion is frequent, due to sulphuric acid, and in those places supplied from wells the usual difficulty is free carbonic acid.
If water works owners could calculate in advance how much monetary damage would be caused to the water works systems in ten, twenty to thirty years, they would undoubtedly apply a remedy, if such were available. The writer has known cases where the capacity of 4 and 6-inch pipes was reduced in twentyfive years as much as 70 per cent., and in many cases where in ten years, the reduction was as much as 20 per cent, for cast iron pipes 12 inches in diameter and upwards.
All waters that cause tuberculation of cast iron pipes are corrosive in action and all such corrosibility is increased enormously with the temperature of the water; for instance, a water which is noticeably corrosive at boiling temperatures, is much more so at the usual boiler temperatures; therefore, if it could be known in advance that water was neutral to iron say at a temperature of 300 Fahr., it is safe to say that it would be neutral at all temperatures below that.
Hence, a simple proposition would he to test the corrosibility of the water at say 300° Fahr., and if found corrosive at that temperature, apply such a remedy as would render the water neutral and free from corrosibility. That proposition sounds simple enough, but in the writer’s experience he has not found to exist any practical application of that principle.
Chemists have concluded that corrosibility is due to ionization or electric action, and have devised methods for determining what is known as the “hydrogen ion concentration.” always using a cold water sample to be tested for this purpose, and elaborate experiments have been made to determine the ratio of corrosibility to the hydrogen ion concentration.
An Effective Method of Testing
Some years ago. the writer had experience with a water supply for a city in the South. Raw water was from a drainage area of about 50 square miles and was high in color, and required over three grains of sulphate of alumina per gallon to bring tlie color down to an acceptable standard. This water, after being treated and filtered, caused considerable red water troubles, and much chemical investigation was made to determine the cause of the corrosibility. Theories were advanced that it might be due to the sulphate of alumina treatment, to free carbonic acid, dissolved oxygen, chlorine or magnesium chloride.
We then devised the following method of testing, which proved to be effective.
A liter of water was placed in a porcelain dish, and a given length of piano wire, weighing about a gram, was submerged in the water, which was boiled for 20 minutes. The water, before boiling, was tested for iron and color, and after boiling a sufficient amount of distilled water was added to bring the sample again up to one liter, and a test for iron and color was again made. The increase of iron in parts per million over that in the unboiled sample was used as an index of the corrosibility of the water. In this manner, tests were made of the raw water, of the water after being treated with sulphate of alumina and filtered, distilled water from artesian wells that was known to be non-corrosible, and also samples from mountain water. We found the raw water had an index of corrosibility, according to this method, of about 2.4. The water after being treated with sulphate of alumina and filtered was about the same. The artesian well water and mountain water and also the distilled water had an index of about 0.60.
Corrosibility Due to Some Carbonic Acid
In order to determine whether the corrosibility was due to carbonic acid or oxygen, these were eliminated by boiling the water before making the test, but as the index of corrosibility remained the same, it was concluded that neither free carbonic acid nor oxygen was responsible, and the corrosibility was due to some organic acid.
By treating the filtered water with various chemicals including hydrate of lime, hydrate of sodium, or carbonate of sodium, we were able to completely neutralize the water. The success of this test suggested a further method, possessing more completeness, to determine the relative corrosibility of water under all conditions, as follows:
A More Complete Test
Samples of water whose relative corrosibility are to be determined, are first analyzed for iron, and then placed in strong two ounce bottles with large mouths and glass stoppers. In each one of these is submerged a given amount of pure iron, a given length and gauge of piano wire in coil being suitable. The bottle is then placed in a pressure clamp and heated to a temperature of 300 Fahr.. which is maintained for say 20 minutes. After being gradually cooled, the water is again tested for iron, the amount of which will indicate with certainty the relative corrosibility of each of the samples of water.
There are two methods of producing this temperature: One is to heat ordinary machine oil in a
pot large enough to hold say ten to twelve samples; the heating may be done by a gas flame or an alcohol lamp. The second method is to place a closed receptacle in a valved by-pass of a steam pipe. This receptacle may be made with a flange cover and have side outlets. Several bottles of water, each with its pressure clamp, can be placed in said receptacle and removed after being cooled, without interfering with the steam supply from the boiler. This method is simple and does not rerjuire any thermometer, because the temperature of the steam in the boiler is generally sufficient to produce at least 300° Fahr.
It is easy to obtain bottles with flat glass stoppers that will not break when heated, and it is important that bottles should be completely filled with water before placing in the receptacle. It would seem to the writer that this method of testing is much simpler, more reliable and superior to any method of determining the hydrogen ion concentration, the conclusions from which are at best conjectural.
Volumetric Test for Iron
The volumetric test for iron is also easy for any one to make with little instruction.
Along the eastern coast of the United States, as for instance, Long Island, and New Jersey, many of the water supplies are taken from wells varying in depth from 40 to 1,000 feet. The water frequently gives trouble due to iron, which formerly was thought to be the principal cause of the difficulty. It has been found in late years that the trouble is generally due to free carbonic acid, which attacks iron receptacles, leaving iron in solution. Occasionally the water is also abnormally low in alkalinity, sometimes giving an acid reaction. It is now believed by many practical water works men that where the alkalinity is less than ten parts per million, the water is likely to be corrosive.
Where the water direct from a well contains at all times more than one part per million of iron, the iron should be removed by aeration and filtration, and if the carbonic acid is also high, the process of aeration will generally reduce that sufficiently.
Use of Nozzles in Removing Carbonic Acid
In many cases, however, the iron content is below a half part per million, and the free carbonic acid content from 15 to 60 parts per million. In those cases, the iron may be disregarded and the carbonic acid removed by several methods; one being to allow it to stand in an open, shallow reservoir for several hours; another being by agitation, and the most effective is to force the water through spray nozzles so as to form practically a mist. Such nozzles as are used in the sprinkling systems for sewage purification are not sufficient for this purpose, but there are quite a number of good nozzles on the market, through which water is forced under from 5 to 7 pounds pressure, that remove at least 80 to 90 per cent, of the CO* content.
Chemically, free carbonic acid can be removed by Determining Corrosive Effects of Water
(Continued on page 997)
(Continued from page 970)
means of lime or sodium, oxide of lime, that is quick lime, preferably in the form of powder, will accomplish the purpose, but as a general rule the powdered hydrate is preferred because it will keep better.
There are two difficulties connected with the use of lime—it takes an enormous quantity of water to dissolve it, ordinarily about 800 parts per million, and even when an excess is used, there is always a considerable amount of residue, and it increases the hardness of the water. Most limes obtainable in the market contain a high percentage of magnesium, which is insoluble, and about the best that can be obtained will have an insoluble residue of at least 10 per cent.
Caustic Soda Useful in Removing CO2
The two sodium compounds usually adopted are sodium hydrate and sodium carbonates; that is, caustic soda and soda ash. For removing CO2 caustic soda has more than double the effectiveness of soda ash, and it readily goes into solution completely, leaving no residue whatever.
This chemical has the great advantage that it increases the alkalinity, softens the water and for complete removal of the free carbonic acid requires 10 parts of pure sodium hydrate for every 11 parts per million of free carbonic acid.
Caustic soda is furnished on the market in three forms: as a massive solid, as a powder, or in the flake form, in all cases in steel drums. It is sometimes an advantage to buy it in liquid form in tank cars or trucks.
Best Way to Determine Amount Required
The best way to determine the amount required is to make a solution, containing one part of flake caustic soda to 800 parts of water, preferably distilled. Then take 100 c. c. of the water to be treated and add 10 drops of phenolphthalein, and shake. The number of c. c.’s required to produce a permanent violet color multiplied by 100, will give approximately the number of pounds per million gallons of flake caustic soda to completely neutralize the CO. In practice it is not necessary for a complete removal, 85 to 90 per cent, being usually sufficient. This is readily allowed for in the test by using 85 c. c. of raw water instead of 100.
At first thought it would appear that caustic soda would be impractical on account of its great expense as compared with lime, which is only 20 per cent, of the price; but the advantage in handling, the elimination of sludge, and not adding to the hardness of the water or increasing the turbidity substantially offset the increased cost. As compared with the spray nozzle system, this usually requires a double pumping and a waste of pressure, which is unnecessary with the chemical treatment. It has also been claimed that the spray nozzle puts an undue amount of free oxygen in the water, which adds to its corrosiveness. Determining Corrosive Effects of Water
(Continued on page 999)
(Continued from page 997)
Formulas for Chemicals Required
The formulas used to determine the amount of chemicals required to neutralize the free carbonic acid are as follows:
With caustic soda—NaOH + CO2 = NaHCO3, sodium bicarbonate.
With sodium carbonate—Na2CO3 + CO2 + H20 = 2 NaHCO3. sodium bicarbonate.
With lime—CaO + CO2 = CaCO3, calcium carbonate.
The relative amounts of each of these three chemicals. as available in the market to use for complete neutralization of free carbonic acid are as follows:
1.4 parts per million of 90 per cent, lime for each part per million of free carbonic acid.
3.11 parts per million of 58 per cent, sodium carbonate for each part per million of free carbonic acid.
0.94 parts per million of 97 per cent, caustic soda for each part per million of free carbonic acid.
The relative costs are as follows:
Per hundred pounds
90% lime ……………………….75
58% sodium carbonate ………….. $1.87
97% caustic soda ………………. 3.65
Taking lime as $1. the relative costs of chemicals in water treatment are as follows:
Lime ………………………… $1.00
Sodium carbonate ……………… 5.65
Caustic soda ………………….. 3.27
So it is seen that caustic soda costs 3.27 times and soda ash 5.65 times as much as lime; but lime hardens the water, is very troublesome to handle, produces a large amount of sludge and sufficient turbidity to require in some cases filtration, all of which are obviated by the use of caustic soda.
Soda Has Same Effiect as Sodium Bicarbonate
As for the effect on the health, soda ash has the advantage, being universally used in cooking and baking in the form of sodium bicarbonate. There are thousands of people who take a level teaspoonful of this in water once a day at the advice of their phvsicians. With the treatment ordinarily used, a man would have to drink 70 gallons of water to obtain a commonly prescribed medical dose, and as an average man seldom takes more than two quarts of water per day. it would take one hundred and forty days to get an ordinary one per day dose.
With the use of sodium chloride, which is common salt, there is much more sodium consumed than could ever occur in water treatment.
It is not necessary to give in detail the methods of practically handling the caustic soda, because conditions differ, and the water works operator has to use his judgment based on the particular situation.
Two Tanks Used Alternately
In one case I have in mind, where the consumption of water is about 1 1/2 million gallons per day, two tanks are in use alternately. Each is about 6 feet in diameter and 4 1/2 feet deep. On the upper portion of each tank is a receptacle, with a screen in the bottom. sufficient to hold about 400 pounds flake caustic soda. A small manifold pipe runs across the top of this receptacle, and this is filled with chemical. Water is sprayed over the top, and at the same time the water is discharged from an opening in the bottom of the tank pointed circumferentially. By the time the tank is full of water, say in a half hour, the entire amount of caustic soda is fully dissolved. This then is fed through an adjustable float valve orifice to the suction well of the main pumps.
It is usually best to make the solution in this manner several hours before it is put into service. The operator soon gets familiar with the routine of the matter, the main point being not to feed fast enough to at any time produce caustic water.
Convenient Arrangement for Feeding Chemicals
In feeding chemicals it is sometimes advisable to take the solution from the upper strata of water, especially where lime is used. For this purpose, Fig. 1 shows a convenient arrangement. However, with sodium hydroxide or caustic soda, this is not necessary, as the strength of the solution is the same from top to bottom, for an hour, if the mixing has been done according to the above description.
Where the consumption of water varies through a wide range, if the solution fed is constant, of course, the amount of applied chemical in parts per million would be subject to the same variation, and this is an undesirable condition. With the device shown, it would be necessary for the operating engineer to keep changing his orifice feed so that it will bear a constant ratio to the consumption of water. This can he done by hand, but requires a good deal of watchfulness and care. There are a number of devices for taking care of this automatically. If the main pumps are of the reciprocating type, there could be made to one of the moving parts an attachment which operates a solution pump, so that the solution would be fed in exact ratio to the speed and discharge of the pump.
There are a number of other devices which are operated by the differential head obtained from an orifice or a venturi tube. One of these is shown in Fig. 2.
To determine whether the water supplied attacks pipes or hot water receptacles, its corrosibility should be tested by the amount of ironthat will be dissolved in water of as high a temperature as will be experienced in practice.
This corrosibility should be neutralized by such methods as are found to be applicable to the particular case, for instance: For coal mine water,
where the corrosibility is due to sulphuric acid, the only satisfactory method is precipitation of the sulphates by barium hydrate… For free carbonic acid
removal, the water can be stored in shallow reservoirs. or there can be used some form of effective agitation, the best being by means of the most efficient spray nozzles. Chemically, oxide of lime, hydrate of lime, sodium carbonate or sodium hydra-te can be used, the most effective of all of these being sodium hydrate or caustic soda.
(Excerpts from paper read before the Pennsylvania Water Works Association at its annual convention held at Atlantic City, N. J.)
Edgewood, Va., Water Works Incorporated—The Edgewood Water Works Company has been incorporated at Edgewood, Va., with J. D. Logan, of Salem, Va., president.