(B) All Meters. (6) The record clerk shall file and have charge of the daily reports made by the meter testers. The yellow sheets shall be kept as records until the books for which they are duplicates shall have been filled and returned by the testers, and then these duplicates shall be destroyed. (7) He shall also make up monthly summaries of tests not later than the fifteenth of the month for the month preceding. For this work he shall use forms G13 to G16, and these forms shall be kept by him on file. WORK OF CHIEF INSPECTOR. (8) All apparatus for the testing of meters shall be examined, approved and stamped by the chief inspector before being used, and he shall also oversee the proving and sealing of all meters. WORK OF COMPLAINT INSPECTOR. (A) Complaint Meters. (9) The complaint inspector to whom an application is assigned shall take the forms sent him, prepare and send out notices when he will call (form G10), and proceed to the premises designated. On arrival there he shall obtain and enter the missing data for the completion of the forms, shall locate the meter, enter its description on the forms and collect the fee, if it has not been paid, entering on the forms the payment of this fee with date and giving the proper receipt for it (form G3) to the consumer. He shall then fill out form G5, detach the white original sheet from the tag and attach the tag to the meter, entering in the proper place on form G4 the date as an indication that the meter has been tagged. At the end of the day he shall send to the record clerk all the copies of form G and all the white sheets of form G5 for the meters to which he has attended, and the receipt stubs (form G3) which he has used. At least once each week he shall account to the clerk for all fees received by him. WORK OF THE METER TESTERS. (A) Complaint Meters. (10) Upon receiving form G4 at the testing station, the tester shall look out for the arrival of the meter described on this form, and when it is received, shall enter in the proper place his name and the date of its receipt. He shall then test the meter, after allowing it to remain in the station not less than five hours in order that the temperature may be equalized, within twenty-four hours after its receipt unless the consumer desires to be present at the test and it cannot be arranged to secure his presence within this time limit. (11) Complaint meters should always be given two tests, and if these two do not agree with each other a third test should be made. If the two tests agree, or if the three tests are made, the average of all the tests shall be taken as representing the condition of the meter. After the test has been made, the tester shall enter the record in the meter test record book, writing "complaint meter" in the "remarks" column and shall also enter upon form G4 the cubic feet passed by the meter and by the prover and the percentage fast or slow shown by the test. He shall also fill out form G6 and paste it upon the meter, and shall seal the meter in the manner hereinafter prescribed, provided the meter is accurate within the limits allowed by statute. At the end of each day he shall send to the record clerk all the copies of form G4 upon which he has recorded the results of tests during the day. (B) All Meters. (12) Before beginning work each morning the tester should see that the temperature of the water in the prover is the same as that of the air in the room, the temperature of the water being adjusted if these temperatures are not the same. For the purpose of observing these temperatures two thermometers must be provided, one being placed in the water and the other hung up in the air, and these two thermometers should be checked against each other to see that when exposed to the same conditions their readings are the same.. The tester must also be satisfied that the meters which he has to test have been in the immediate vicinity of the provers for a sufficient length of time to have acquired the temperature of the air surrounding the provers. While testing, no direct sunlight should be allowed to strike either the provers or the meters to be tested and the observations of the temperature of the water and of the air shall be made at least every fifteen minutes. (13) Meters are to be given the check test only, but while making this test the size of the dial is to be observed, and it is to be noted whether this dial is of the proper size for the size of the meter. (14) After the meter has been connected to the prover, a test for the tightness of the connections must be made by closing the meter outlet with the hand and throwing the prover cock wide open, so that full pressure is thrown on all the connections and the meter, and then shutting the prover cock and watching the pressure gauge to see that the pressure holds. If the pressure does not hold, the connections must be gone over and another test for tightness made before the test of the meter is begun. (15) From 1 cubic foot to 1% cubic feet of air must be passed through the meter before beginning the test. The object of this is to make certain that the meter is working freely before the test begins. (16) The proving hand on the dial must be brought to, and stopped at, one of the divisions on the up stroke of this dial and the test started with the hand at such a division and concluded when the hand again reaches this division after making a complete revolution. (17) Meters showing on the first test a proof of either 99 or 101, or one between these limits, are to be passed on this single test. (18) Meters showing on the first test a proof between 98% and 99, or between 101 and 1012, or exactly 98% or 1012, are to be given a second test, the average of the prover readings and of the percentages of errors for the two tests being recorded and the meter being passed, provided the average proof is more than 98 and not more than 102. (19) Meters showing on the first test a proof between 98 and 98%, or between 101% and 102, or exactly 102, are to be given a second test, passing four cubic feet through three and five-light meters on this second test, and are to be passed if the average proof shown by the two tests is above 98 and not more than 102, the average of the prover readings reduced to the normal 2 feet basis for three and five-light meters, and the average of the percentages of error for the two tests being recorded. If necessary, a third test may be made with such meters, but only when there is some doubt as to the accuracy of the first two tests. If the third test is made the average of all three tests is to be taken as being correct, and the meter is to be passed or rejected on this average. (20) Meters showing a proof of 98, or less than 98, or more than 102 on the first test, or as an average of either two or three tests, as specified above, are to be rejected and returned for readjustment by the company presenting them for test. Form G6 shall be filled out and attached to each meter rejected. After such readjustment the meters are to be again tested. (21) The bell of the meter prover should be drawn to its full height and the scales wiped the last thing at the end of each day's work and left up during the night. The object of this is to prevent discoloration of the body of the bell and of the scales. Care should be taken that the cocks are shut off tight so that the bell will remain up all night, and by taking the reading on the scale at the time of stopping work and again when work is begun the next morning a test is obtained of the tightness of the prover and the connections up to the prover cock and of this cock itself. In case these readings show that the prover is not tight the leak should be found and remedied before making any tests. (22) The record of every meter tested should be recorded in full in the meter test record book, and in the "remarks" column a note shall be made to show whether the meter is a new, repaired, complaint or removed meter, wherever possible. (23) Every meter ascertained to be correct within the limits allowed by law shall be sealed, and no meter shall be sealed which has been found to be incorrect according to the preceding instructions. The tester should see that the ring is properly placed upon the meter, so that the ring will overlap the seam on the cover. In no case shall a meter be sealed unless the ring is properly placed. Sufficient wax should be poured into the ring so as to permit the making of a clear impression of the seal, and the impression should be clear and easily legible. A table showing the percentages of errors under varying conditions appears as an appendix to this report. Respectfully submitted, TABLES SHOWING PERCENTAGE OF ERRORS IN METERS WHOSE REGISTRATION DIFFERS FROM THE INDICATIONS OF THE TEST GAS HOLDERS. MR. FORSTALL'S REPORT. Alfred E. Forstall. Consulting Engineer,] New York, July 27, 1907. Public Service Commission First District, State of New York, No. 320 Broadway, New York City: GENTLEMEN-I beg to submit the following report upon the manner in which new and repaired gas meters should be tested and sealed, and also upon the procedure to be followed with regard to applications from gas consumers for an official test of their meters: TESTS TO BE MADE. The official testing of gas meters has, in every case with which I am acquainted except two, been confined to testing for accuracy of registration in passing gas at a normal rate. The two exceptions are those of Great Britain, where the Sales of Gas Act, 1859, which governs this testing, provides for a test for "soundness or leakage to determine whether or not the meter registers a small consumption; and of Minneapolis, where both a test for registration under small consumption and one for regularity of delivery under a large consumption are provided for in addition to the tests for accuracy of registration when passing gas at a normal rate. In my opinion these exceptions to the general rule cover matters that are entirely without the sphere of an official test of gas meters. The only object of such a test is to give the gas consumer an assurance that the instrument employed for measuring the amount of gas that he consumes gives correct measure, and this is fully covered by the test for accuracy of registration when the meter is passing gas at a normal rate. A meter that will not register under a small consumption will register slow under normal conditions, and if the fault is sufficiently serious it will cause the meter to be rejected when tested for accuracy of registration. Irregularity of delivery under a large consumption does not affect the accuracy of registration, and if the gas which is being measured is to be used for fuel purposes it does not even affect the quality of the service unless the irregularity is very pronounced. It therefore is not a matter with which the official testing of the accuracy of the meter as a measure has to do. I would therefore advise that the only test of a gas meter made by you he that for accuracy of registration when the meter is passing gas at the rate of six cubic feet per hour for each "light" of its size. METHOD OF MAKING TEST. This test, which is known technically as the "check " test, is made by connecting the meter, with its outlet capped by a check, in which is a circular orifice of such a size that when the pressure on the meter is adjusted to 15-10 inch of water air will pass through this orifice, and consequently through the meter, at the rate named above, to a "meter prover" and passing from the prover through the meter the amount of air required to cause the proving hand on the meter dial to make one complete revolution. This revolution corresponds to the registration on the meter dial of a definite quantity of gas, which varies with the size of the meter. The "meter prover," which consists of a holder, that is, an inverted sheet metal bell, guided so as to be capable of moving freely up and down in a sheet metal tank which is filled with water to seal the otherwise open bottom of the holder, carries on the side of the holder a scale graduated in feet and tenths of a foot over its whole length and in hundredths of a foot over a certain portion of its length. The difference between the readings of this scale at the beginning and end of the test shows the volume of air that has actually passed out of the holder and through the meter. If this volume corresponds exactly to that registered by one complete revolution of the proving hand of the meter the latter is correct, while if it is larger the meter is low, and if it is smaller the meter is fast, the percentage of error in either case being found by dividing the difference between the prover reading and the meter reading by the prover reading. Thus If the prover reading is 2.02 cubic feet, while that of the meter is 2 cubic feet, the meter is 2:820.99 per cent. slow, while if the prover reading is 1.98 cubic feet, while that of the meter is 2 cubic feet, the meter is 1.01 per cent. fast. The accuracy of the graduation of the scale on the meter prover holder is assured by having the graduations marked in the first instance as the result of a careful calibration of the holder by means of a cubic foot bottle certified by the National Bureau of Standards at Washington and by periodical testing by means of such a cubic foot bottle in order to provide against any error arising through deformation of the holder. The prover holders are made with working capacities varying from two cubic feet up to ten cubic feet, and sometimes are made even larger, for use with different sized meters, the five-cubic-foot size being the most common. Before making a test It is necessary to have the temperature of the water in the tank of the prover the same as that of the air in the room and that of the meter to be tested. To insure the latter condition the meter must be allowed to stand in the testing room at least five hours before being tested and care must be taken to keep both the prover and the meters to be tested out of the direct rays of the sun. These and other precautions that should be taken to insure the accuracy of the test are covered in the Instructions for Meter Testers appended to this report, |