METERING OF PRIVATE FIRE LINES

Test of Devices Made at Lowell.

Fire service meters were discussed at the recent convention of the New England Waterworks association in New York. Mr. E. V. French introduced the subject as follows:

Several years ago after a number of discussions in this association and elsewhere of the problems of the age, “The private fire service,” we came to the conclusion that the best solution for our difficulties would be a meter which would assure all the services, but could not seriously obstruct the flow. There was no meter on the market fulfilling these requirements, and we started to devise a meter for this use. We hit upon the idea of getting the action of the bypass meter. The working out of this idea has proved to be one way of making a meter that will assure small, as well as large flows. We have discussed the problem of the Hersey meter, and have taken up the matter in the various opportunities that presented themselves to us. We have kept clearly in mind that the chief thing was to have on the market a safe and reliable meter. The Hersey Detector seems to us to be valuable in all cases for which it was designed and not objectionable to any insurance company. We stated that we would not object to the Hersey device, and we will not object to any similar device performing the same duty in introducing hazardous features. Our present position in regard to fire service is as follows: (t) We make the mill people cut off any connection of water pipes for manufacturing purposes. (2) We put on a Detector meter, such as is made by the Hersey people, so that the mill forces clearly understand that they must not draw water from the fire service pipes, except in case of fire. (3) If the water denartment feels that a meter assuring all flows is necessary, we hope that a meter such as that described will soon be available. We believe that today it is feasible to put a guard over the fire service, which, with reasonable care, will give the water department full knowledge of what is going on. The National Fire Protection association has appointed a special committee of five, including myself, to consider the whole fire service problem. This association met a committee of the American Water Works association last year and discussed problems with its members, and it may he interesting to quote a vote that had been passed to show the attitude of these people. “It is the desire of this committee of the National Fire Protection association to discuss these matters in a friendly way with the waterworks men, and it is believed we need the advice of the men handling the water supplies. This committee is, therefore, ready to discuss with the committee of this association any problem, and we believe there will be no question which cannot in this way be satisfactorily overcome.”

Mr. Thomas said: As to the idea I had of inspecting the meter referred to at Lowell last Saturday: 1 was sorry we had no time to make special tests of it: but with the tests that had been previously made I was somewhat familiar, and 1 was thoroughly satisfied that the meter would perform the duty for which it was designed. It seems to me Mr. French has gone nearer to what is required along this line than has been accomplished heretofore. He certainly has given a meter which is acceptable to the insurance comoanies, and that is saying a great deal. As we all know who have used them, the threeinch disk meters will measure streams fine enough to satisfy even the most hypercritical waterworks man. Fire service which was metered had an advantage, as it showed that the apparatus was in working order. If the owners of mills can be induced to put in three or four connections instead of one, I am sure the water departments will have no objection. 1 am glad to see the result that has been accomplished by the association to which Mr. French belongs. Those of us who have heard him discuss this question before undoubtedly recall his various remarks. I think this whole question of stealing water will come very close to the end where political matters do not interfere. This association, and the American Water Works association can take a great deal of credit to themselves in enforcing the accomplishment of this end.

Mr: Kimball said: There has been a very earnest effort to adopt some kind of a meter device meeting the requirements of the underwriter and of the waterworks man. The requirements are very severe. The underwriter, on the one hand, says that you must not put anything on a main fire line which will in any degree obstruct or endanger the flow. In the attempt to work out this problem, we went to the underwriters and asked what they would allow. We must have some device on the pipe, in order to undertake to do anything in the way of measuring water, and, when that was boiled down, the only thing seemed to be to taxe a self-acting chain as Mr. French uses in his meter. The waterworks man wants to get all flows down to a gallon or two a minute, and anything from that clear up to the highest rate, bearing in mind that, if water is used for fire protection on these lines, it must be used legitimately. Mr. French, with his infinite patience and skill, has gone still farther than we have. Our device and the French device are parallel up to a certain point. If as a matter of record it is desired to parallel the French device, 1 will submit Mr. French’s report of the Hersey Detector meter as prepared by him as chairman of the private service committee of the National Fire Protection association committee.

Mr. French said: I wish to pay a compliment to Mr. Kimball, when, as 1 suppose, he desires to measure water down to one drop a week. I should like to know the general feeling of the waterworks men as to the need of a meter measuring all flows. I should like to know whether in the judgment of waterworks men a similar detecting device which would measure all flows up to 150 gallons a minute would not in a great many cases be all that was necessary.

Mr. Sullivan said: At the tests at Lowell the meters were neither accepted nor rejected. I was one of the fortunate ones that saw the meter that Mr. French spoke of today. Therefore, the paper which you have before you is identical with the paper of March last. As regards the welfare of the community: It strikes me that the waterworks managers have the welfare of the community at heart as well as this body of men. It is to be believed that waterworks managers can agree on some mechanism that will meet with the approval of all. I do not believe it was the intention of the Lowell water department to make a series of fire tests. We made a series of experimental tests only, and approved one in which, in each case, we had a high-service pump working on a relief-valve. Mr. Nash said: I can not add anything to the discussion for the reason tha’t I entered into this matter a little late. Some years ago 1 made an experiment in which was used a great number of proportional meters, and in one of these devices a valve was used to govern the flow of the water, and the device was expected to register from the ground, so as to give a proportional amount of water in proportion to the opening of the valve. The invention was laid on the shelf, because at that time there was no device needed for this special fire service. Of course, the success of a device will depend on the mechanism as conducted. Mr. Thomas said: In answering the question of Mr. French in a great many cases, the Detector will be perfectly suitable; in other cases it will not. This matter of metering fire services has been agitated for many years. Owing to the absence of Mr. Crandall, we have not made a full report, but the paper of Mr. French is an intimation of the good work that has been started by this association and carried along by’ this committee. The papers read by Mr. Sullivan last March and that by Mr. Kimball were indirectly the work of the committee. The fact that water through fire service lines should be metered was recognised by the members at our previous discussions. The American Water Works association voted in favor of the meter, and there seems to be quite a unanimous opinion that fire services should be metered. Another matter in regard to fire services, it is well understood, is that most municipalities are willing to defray the expense of putting in pipe lines in mill and breweryyards for fire services, so that it is not intended especially in the name of any industry, and the owners of these places will use the lines, if they can, without paying anything for that use. The cities are willing to give any amount of water to extinguish fire; but, of course, the laying of mains to accommodate large concerns is a privilege which should not be abused. We have a concern in Lowell that has two fire services; two six-inch and two eight-inch mains. We expect to have every fire service in the city metered, and it is the duty of our waterworks men to interest themselves in the way water is being wasted, and, if to a great extent without authority, a meter should be put in. Mr. French said: We expect to have every fire service in cities metered and the Hersey device happens to be the only one now available from long experience. We would not insist on metering everything that we see right away; but we would like the progress to go slow, so that we might find out difficulties. It is one of the cardinal principles that very large pipe going into a building should have an outside shutter. We invariably put a gate on every building and locate it outside a building, placing it so that it is always accessible, even if the building is not on fire. While the private fire service does put a large pipe in the building, it at the same time immensely reduces the chance of a fire in that building and makes that building a strong barrier from conflagration. There is no better way to prevent fire than equipping buildings with automatic sprinklers, and these buildings have to be provided with a great deal of water. Mr. Sherrerd, of Newark, N. J., said: I appreciate that it is an advantage to have these large buildings properly protected; but a great many representatives of the insurance people, I am sorry to say, impress me with tbe fact that they are only looking after the particular risk on which they are working. Mr. Holly said: The paper, as reported this morning regarding the charging for water filtered and the point we have reached, presents the matter to the attention of the association which should be taken up, and that it is the proper basis on which to estimate the value of fire protection to manufacturing concerns that have special protection. ,

On motion it was resolved that a committee of five should be appointed to consider the matter and report at some future meeting of tbe association.

Lowell, Mass., having determined on metering every one of its 124 fire service mains, twelveinch to one-inch, instead of only one twelve-inch, determined on making tests of meters as detectors of the flow. The water board made the tests with whatever facilities were at hand, and, as sucn tests would be incomolete and inconclusive, decided to vary the usual test of separate meters by connecting them in series and getting the comparative results. These are given in detail by Assistant Engineer William F. Sullivan in a paper on “Tests of large meters and fire service devices,” read before the recent convention of the New England Waterworks association, and are herewitn presented in abridged form : The meters were set on the basement floor and near the small testing rig. The water for the tests was supplied through a four-inclt pipe used for fire service and testing purposes, and fed by a sixinch street main, about twenty-five feet from the building, the street main in turn being supplied from a twelve-inch distribution pipe direct from the low-service reservoir, insuring a constant and steady supply. The first arrangement, was made up of three current meters, a six-inch Crest, sixinch Gem and six-inch Torrent, forming the letter U. On the outlet-side was placed a six-inch Tilden’s device with three-inch Hersey disk meter on bypass; from the outlet end of this device a fourinch riser was run to the testing apparatus for small meters. On this riser was tapped a connection for supplying the shop through a threequarter-inch Empire meter with water for ordinary purposes, such as the intermittent use of water for two sinks, two waterclosets, and one washstand. During tne time tbe shop was getting its supply from this source, the regular service supply was cut off, and the water used for testing small meters at the weighing tank was checked and recorded by a one-inch Crown meter. As the test proceeded, it was found that the threeinch Hersey disk was a close registering meter and agreed with the amount of water recorded by both the three-quarter-inch Empire and oneinch Crown. While testing the current meters for sensitiveness and accuracy on different sized streams tnrough the multiple cock and weighing the discharge in the tank, the shop was supplied by its own service. Under the “ordinary use” test, the board undertook to find which of the current meters would show the largest percentage of registration under conditions similar to those met with in practice, where water is used, or wasted through carelessness or otherwise. On testing for sensitiveness on one-sixteentn-mch to one-half-jnch streams, requiring considerable time for the registering hands to make complete revolutions on the dial, so as to eliminate any error due to the adjustment of the index hands, the apparatus could be set in operation and allowed-to run unattended for hours or days, and, by careful timing, gauging of discharges before and after each test, and also comparing the check meters, very accurate results were obtained. Before beginning the series test the meters were thoroughly flushed through a two-inch opening. The first meter on the line failed to register, and an inspection showed that several ounces of soft, slimy sediment had accumulated on the working parts of the meter; but tuts was considered insufficient to derange in any way the proper working of this type of meter. Shortly before the meters were flushed, a half-inch hole was drilled and tapped on the inlet side of meter to connect the pressure gauge, care being used that none of the drillings should get into the meter. None were found at the time the meter was opened and examined. This failure to record brought up the question: Did position in the arrangement or line have any effect on registration? Some meter men, owing to a lack of better information, were inclined to believe it to be so. To satisfy some of the meter men and gather information on this point, a new arrangement and alignment of the meters was made, the meters being made interchangeable by fitting with nipples and uniform flanges. The meters were then successfully tested under various conditions. During a portion of the test under one condition, it was found that small rates of’flow through one of the current meters did hot cause any movement of the register until the one-inch stream, with sufficient water pressure to overcome the adhesion, set it in motion. They tried repeatedly afterwards to obtain like results on small streams; but on each occasion the meter registered. After obtaining considerable practical information with, the facilities limited to a two-inch opening on the multiple cock, inadequate drainage for larger flows and a small weighing tank, they fully realised that to make more complete tests some other arrangement would be necessary. In studying the best methods of measuring the flow of water and checking the registration of meters, the nozzle seemed to ^ be best adapted for such work and could readily be employed as a practical check meter; they were fully convinced that Mr, John R. Freeman’s experiments with nozzles were conclusive. As Mr. Freeman says, “The nozzle will, I think, on reflection, be granted to be the most portable and compact gauging apparatus in. pro-, portion to its capacity which has ever been devised for such purpose.” To give further confidence in the use of the nozzle, Mr. Freeman asked himself the question, “Is any accidental variation in method of setting tip liable to introduce a change of rate or error in its indication while using the coefficient as previously determined once for all?” In answer to his own question, he says, “In the subsequent use of this particular piece of apparatus, we may, (if our pressure gauge is all right and the apparatus set up in almost any reasonable manner) have the greatest confidence that the error of any series of measurements would not exceed one-half of one per cent.” t he expense in such installation is not prohibitive. To establish a large meter-testing plant, the necessary requisites are, first, an abundant water supply; second, readiness for raoid drainage; third, suitable apparatus. The testing station at Lowell was in the engine-room of pumping station 1. It had nearly all the requisites, with the additional advantages of being ready for use at any or all times, economy of water, compactness, light, heat, pleasant and convenient surroundings, and proximity to waterworks repair shop. This station is located about one-half mile from the lowservice reservoir, with a capacity of 30,000,000 gallons, supplying the city proper through a thirtyinch delivery pipe: branching from this pipe, near the reservoir, a twelve-inch distributionpipe line runs directly down hill and into the basement of the pumping station, there being reduced to an eight-inch pipe, supplying the highservice pump. An eight-inch branch was cut into this line and the pipe laid under the basement floor to the westerly wall of the station; an eight-inch elbow, with riser extended to twentyfive inches above engine-room floor; another eight-inch elbow reducing to six-inch pipe, with a six-inch gate-valve controling the inlet. From this valve about eight feet of six-inch wroughtiron pipe to inlet piezometer; next a space for the insertion of any size meter or device from three inches to twelve inches, the meters being swung into position by means of a hoisting rigging; on the outlet flange of the meter another piezometer ring was attached, then about eight feet more of wrought-iron pipe to pressure piezometer and nozzles; between the nozzles and meter a six-inch valve was sometimes inserted to keep the apparatus from draining and at the same time obtain conditions similar to what is often found in actual use; the nozzles discharged through a sheet-iron spatter box, seven and one-half feet long, into and against a twenty-four-inch castiron deflector elbow set directly over the pump well; this deflector complies with the second requisite for testing, readiness for rapid drainage, without discharging or wasting into the open. The whole apparatus was held up in position with adjustable braced wooden chairs, with iron clamps holding the pipes firmly in position without undue vibration. Under the whole string of pipe and apparatus is placed a drainer pan to catch the condenstion, drippings or bleeding of the line, all of which wastes into a sewer drain. To obtain nozzle discharges, the methods used were substantially those given by Mr. John R. Freeman in his paper on “Hydraulics of lire streams,” and those in his paper on “The nozzle as an accurate water meter,” and from methods described by Mr, Frank C. Kimball in his paper entitled, “Some six-inch meter tests and how they were made.” The nozzles used ranged in size of orifice from one-eighth inch to four inches and were made by the Hersey Manuacturing company, and constructed in accordance with the requirements from experiments by Mr. Freeman. The tests for small flows—that is, one-sixteenth-inch, oneeighth-inch and one-quarter-inch streams, were determined by attaching a two-inch gate-valve on end of two-inch nozzle and screwing ott outlet side a brass cap, with circular brass disk, having a “standard circular orifice with sharp edges.” The coefficients for these discharges were determined by experiment. The water pressure was measured by means of an open mercury column, with graduated scale reading to a tenth of a pound. The mercury pot was one of the regular pattern, having a small glass window in the side permitting a comparison of the zero point of scale with mercury in the pot. For measuring the retardation or loss of pressure in the meters, a mercury U-gauge, placed in a cabinet and fastened to the wall, was used. To facilitate the work of testing, a complete collection of adapters was procured, including different lengths of six-inch pipe and nipples, flanges, with standard drilling, reducers, expanders, valves, etc. By the use of these adapters any machine from three inches to twelve inches could, in a short time lie inserted in the apparatus ready for testing. The elevation of the centre of the nozzles was 39.6 feet above city datum; the level of high water in the reservoir is 181.5 feet, thus giving a static head of 141.9 feet, or a pressure of about 61.4 pounds. Notwithstanding that the twelve-inch pipe from the reservoir fed several laterals of the supply system, the pressures given by the mercury column were reasonably free from fluctuations. Assistant Engineer Sullivan, Assistant Engineer and Superintendent R, J. Thomas, of Lowell, Mass., give the following on the subject of special measuring devices for private fire service lines tested by them. The results of some current meters tested were as follows:

Six-inch Hersey Torrent—On flows under fiftygallons per minute registration is variable; between one-half-inch and four-inch streams the average is ninety-nine per cent.

Six-inch meter (Neptune Meter company)— Eighty-nine per cent on a flow of nine and onehalf gallons per minute; on higher flows from ninety-two to 103 per cent.

Six-inch meter (National Meter company) — On flows above fifty gallons, 99.12 per cent; below thirty gallons, correct.

Ten-inch Hersey Torrent—Flow’s from fiftysix to 2,410 gallons g6.i per cent.

Twelve-inch Giant Hersey Meter—On flows over one-half inch 94.8 per cent.

The six-inch Tilden device showed an ability to register within four and one-half per cent of the available supply.

The conclusion of the report says: The figures show that some of the special devices can be depended upon to deliver more water with less loss of pressure than either the positive or current type of meter; that the devices are capable of recording the flows for which they are adjusted, but beyond such adjustment the flows can only be approximately accounted for. If metering appliances were placed on all private supplies it would effectually put a stop to chronic wastes, and there would no longer be any need for the department to supervise and inspect the fire service beyond such device, placing upon the parties protected by these fire services the responsibility of ascertaining the inherent leaks of the hidden portions of the fire-service system; for, undoubtedly, the party having to pay the bills will be inclined to stop such wastes, even to the extent of relaying and renewing the underground furniture and fixtures.

The following is a description of the Hersey Detector meter for fire service: The object of the Hersey Detector meter is to measure accurately drafts up to about 150 gallons per minute, and to detect and exhibit the flow of any additional quantity in excess of the above amount. The device consists of an increasing check-valve in the fire service main and a meter on a bypass round it. There is no mechanism in the main pipe except the check-valve, which offers no objectionable resistance, and gives a practically unrestricted water-way for use in case of fire. What the detector does, is, first to cause all drafts up to about 150 gallons per minute to go round through the bypass meter, where they are all measured; second, to give a positive indication that, while this is going on. the check is closed; third, to give a positive indication if the draft exceeds 150 gallons per minute, and about how many hours such an excess has been going on, and. if desired, to indicate about when it took place. The device is furnished completely assembled, so that it may be set as one piece in the main pipe line. It may be assembled, so that the bypass will lie on either side of the main fire line, or on top or underneath, if more convenient. The bypass is controled by valves so as to permit inspection of the meter without closing the fire line. In case it is desired to use the whole device in place of the main yard check, the valve on the mill side can be a check-valve. The main check-valve is provided with a differential seat, which, with the bypass, gives an initial resistance of about six per cent of the available pressure. When the bypass meter is delivering about 150 gallons per minute, the friction loss through the meter reduces back pressure on the check to the equalising point and any draft, in excess of 150 gallons per minute, unseats the check and finds unrestricted passage through it. The moment the check leaves its seat, a small amount of water nows from the differential seat to the atmosphere through the small indicating meter, and this meter will continue to run as long as the check is off the seat. I he check is so weighted that it will close when the flow drops somewhat below the flow which caused it to open. The device will, therefore, accurately measure all leaks or ordinary small drafts, will give unrestricted water-way for use in case of fire, and will detect and show if water has been used in very large quantities, giving an approximate idea of the length of such use. and of about when such drafts occurred. In finding the relation of difference between the meters and devices tested and straight plain pipe, the space by the meter in the apparatus was fitted with a three-foot “filling-in piece” of six-inch pipe. Under this condition the quantity of water discharged and loss of head were obtained by the use of different sized nozzles, thus showing the available delivery and loss of pressure or friction in the apparatus as follows:

With tests under similar conditions, the above table permits a comparison of flows and loss of pressure between the meters and devices tested and the maximum discharge and minimum loss of pressure. The only correction necessary, and that is inappreciable, is for the difference in static head due to the slightly varying level of water in the reservoir.