1. Definitions [1, 2]1

absolute pressure. Total pressure measured with respect to zero pressure [i.e. psia or kPa (absolute)].

atmospheric pressure (mean). The atmospheric pressure agreed to exist at the meter at various ranges of elevation, irrespective of variations in atmospheric pressure from time to time.

badge. A metal plate affixed to the meter by the manufacturer showing the manufacturer's name, serial number and model number of the meter, its rated capacity, and other basic meter information.

base pressure. The absolute pressure used in defining the gas measurement unit to be used, the gage pressure at the meter plus the mean atmospheric pressure. For most practical purposes the mean atmospheric pressure is the barometric pressure given in the altitude correction tables (table 1). For example: if the gage pressure at the meter is 11 in water column (WC) and the agreed atmospheric pressure is 14.4 psia, then the base pressure would be 14.4+11/27.7= 14.4+0.4 14.8 psia. There are 27.7 in WC in one psia.

bell prover. A calibrated cylindrical metal bell with a scale thereon which, in the downward travel in a surrounding annular tank containing a sealing medium, displaces air through the meter being proved or calibrated. The prover may be calibrated by "strapping" ( a measurement and calculation technique) or by "bottling" (a volume comparison technique) with a standard cubic

'Figures in brackets indicate the literature references on page 17.

foot bottle or a Stillman portable cubic foot standard.

capacity rate. The rate of flow (in cubic feet or cubic meters per hour) of a liquefied petroleum gas vapor-measuring device as recommended by the manufacturer. This rate of flow should cause a pressure drop across the meter of 1/2-inch WC.

check rate. A rate of flow usually 20 to 35 percent of the capacity rate.

and

cubic-foot bottle. A specially constructed calibrated metal bottle open at the lower end and so supported that it may be easily raised or lowered in a tank which contains a sealing medium. With the level of the sealing medium properly adjusted, the bottle, when lowered, will displace exactly 1 cubic foot of air upon coming to rest on the bottom of the tank. The marks on the bottle defining the cubic foot are the bottom of the lower neck and the gage mark which partially surrounds the gage glass in the upper neck. The calibration of the bottle must be traceable to the National Bureau of Standards.

cubic foot, metered. The quantity of gas that occupies 1 cubic foot when under pressure and temperature conditions existing in the meter.

cubic foot, standard. The quantity of gas defined in gas industry as that quantity which under a pressure of 14.73 psia and at a temperature of 60 °F occupies a volume of 1 cubic foot. (A quantity defined in the gas industry as that amount of gas that occupies 1 cubic foot under the conditions indicated.)

cubic meter, metered. The quantity of gas that occupies 1 cubic meter when under pressure and temperature conditions existing in the meter.

pilot flow rate. A minimum flow rate that a meter is required to register with a prescribed accuracy. pressure drop. The loss in pressure between two points in a fluid flow system.

prover oil. A light oil of low vapor pressure used as a sealing medium in bell provers, cubic-foot bottles, and portable cubic-foot standards.

proving indicator. The test hand or pointer of the proving or leak-test circle on the meter register or index used for testing the meter and for indicating gas flow.

Stillman portable cubic-foot standard. A gasometer of the annular type, the bell being sealed with a light oil, the amount of its rise (and consequently of the volume of air or gas being measured) being under absolute control so that an exact cubic foot can be delivered.

strapping. A method of checking a bell prover by determining the relation between displaced volume and linear movement of a bell prover by

means of measuring scale length, bell circumference and displacement of the sealing liquid. vapor-measuring device (liquefied petroleum gas). A system including a metering mechanism or device of the meter type, equipped with a totalizing index, designed to measure and deliver liquefied petroleum gas in the vapor state by definite volumes, and generally installed in a permanent location. The meters are similar in construction and operation to the natural and manufactured gas meters. The majority of the meters are of the conventional diaphragm-type.

2. Testing Methods

The customary basic method of testing a LPG vapor meter for accuracy of registration is by passing a measured volume of air from a bell prover through the meter and comparing the registration of the meter with the volume indicated by the prover.

The use of reference meters for testing the accuracy of LPG vapor meters has certain limitations since a prover is still necessary to frequently verify the accuracy of the reference meter and an additional uncertainty due to the use of a prover is added to the test method, necessitating an increase in tolerances.

3. Testing Apparatus - Bell Prover

The testing apparatus (description, installation, calibration, and maintenance), prover room, and basic test procedures are well described by Beck [3]. Collett [4] also has discussed the calibration of bell provers. Both have given permission to draw freely from their publications.

3.1. Description

The tank and bell portion of a typical bell prover is shown schematically in figure 1 in both the raised and lowered positions. The bell prover consists of an annular metal tank (prover tank) open at the top and nearly filled with a sealing medium (usually an oil), in which a cylindrical tank called the bell, open at the bottom and having a dome-shaped top, can be raised or lowered. The liquid acting as a seal prevents air from entering or leaving the bell except through a pipe in a dry well which extends from the outside of the prover tank down underneath it and up on the inside to a point above the level of the liquid. The prover tank is made of two concentric tanks with a distance of about two inches between them. The inner tank has an airtight top through which the air supply pipe passes. This method of construction reduces the amount of oil required to seal the bell and when raised off the floor by the prover base permits more rapid temperature adjustment of the oil.

Figure 2 illustrates the exterior of a bell prover. The bell is guided in its upward or downward movement by sets of rollers, at the top and at the bottom of the bell. These guide rollers revolve against the guide rods. The frame is supported by posts or columns which are fastened to the top rim of the prover tank. Two sets of antifriction rollers, which support the axle of the large counterbalance wheel and the cycloidal lever arm, rotate in brackets on the top of the cast iron frame. The weights which counterbalance the prover bell are attached to a Ichain which is carried over the counterbalance wheel and fastened to short chains or cables which are equally spaced and attached to the edge of the dome of the bell.

When air is removed from the prover and the bell descends into the liquid, the pressure would decrease, due to the increase of the volume of the submerged portion of the bell, if it were not for the shape of the cycloidal lever arm. This is designed so that as the prover bell descends into the liquid, the cord attached to the arm and supporting the secondary counterbalance weight moves in toward the axle decreasing the counterbalancing action of the smaller weight just enough to compensate for the loss in pressure due to the increased immersion of the bell.

At the top of the dry well or air pipe is a circular slide valve through which air is admitted to the prover, and at the side is a quick acting valve through which air is passed from the prover to the meter. A siphon-gage or U-gage is connected to the piping between the outlet valve and the meter for use in testing for leaks in the meter or connections to the meter.

Prover tanks are usually constructed of brass and bells of copper or stainless steel. They are usually made in capacities of 2, 5, 10, and 20 cubic feet.

Special provers of 50, 100, and 200 cubic feet capacity and larger have been manufactured. Provers are regularly supplied with hose and fittings for connection to the meter, speed unions, and thermometers for air and liquid temperatures. It is desirable that the volume of the prover should be equal to or greater than the full reading of the proving circle of the index of the meter (see fig. 9).

Prover scales have generally been in whole and decimal units of the cubic foot. However, since meters for domestic use are already being made in whole and decimal units of the cubic meter, consideration should be given to the use of a combination prover scale in units of cubic feet and cubic meters. Prover builders have indicated that it would be no problem to supply such a combination scale.

Bell-type meter provers are normally used for pressure proving, that is, with the provers loaded to a pressure about 1.5 in of water above atmospheric pressure. To change the prover pressure, add or remove some of the main counterbalancing weight (additional weights placed upon the crown of the bell should be avoided). After the prover pressure is set, change the amount of sealing liquid until the

The prover should be checked daily for leaks. A leak test may be made after the temperatures of the room air and the liquid have equalized. Raise the prover bell and set the marker at zero on the scale or read out device. If there is any noticable change in indication in a 10-minute period, determine the cause and make the necessary corrections prior to resumption of testing. A more comprehensive leak test may be initiated in the same manner as the above procedure by permitting the bell to remain in this position for a period of 6 hours or longer (overnight). If the rate of change in reading is greater than 1 percent of the rated prover volume per 6 hours [5] determine the cause and make corrections before using the prover for testing meters. It is possible that the prover may not remain at zero, as expansion or contraction due to temperature changes and barometric pressure changes overnight may cause a rise or fall of as much as an inch or more on the scale. This expansion or contraction should, of course, be ignored. If on any particular day the change exceeds this amount, investigate for possible leaks.

The prover should be balanced in such a manner that the pressure of the air in the bell is constant regardless of the position of the bell. To verify this, the prover bell should not rise or fall at either the highest or lowest position of the bell when the prover valve is open and the prover pressure reduced to zero. If this is not the case, the counterweight of the cycloid should be adjusted. This is accomplished by adding or subtracting weight from the bell counterweight when the bell is in the lowest position and adding or subtracting weight from the counterweight of the cycloid when the bell is in the highest position. This is repeated until the bell will remain stationary at any point when the valve is open. The main counterweight may then be decreased until the desired prover pressure is obtained.

Originally water was used as a sealant in bell provers and cubic foot bottles. However, in view of the errors introduced into measurements of volumes of air in contact with water, it is highly advantageous to use a light oil of low vapor pressure instead of water. The advantages of this are:

1. Avoidance of errors due to change in volume by the introduction of a vapor into the air to be measured.

2. Elimination of errors due to a change of volume caused by the cooling effect of evaporation on the outside surfaces of the bottle and the proving bell.

3. The temperature of the prover will follow changes of room temperature more closely due to the low specific heat of the oil.

4. The testing apparatus is preserved in good condition by the sealing liquid.

Thus, the cubic-foot bottle and the meter prover may be compared with greater accuracy than when water is used, and with much less trouble. Also, the errors in actual meter testing are reduced since there is no undesirable change of volume of the air-assuming the temperature of the room is

The operator should keep the rollers and roller guides clean and well oiled. He should test the prover each week or month to see (a) that the bell will drop when loaded to a pressure of 0.05 in of water, or (b) that the variation in prover pressure does not exceed 0.05 in of water during the drop of the bell. The former test is made by raising the bell to the highest position, adjusting the prover pressure to 0.05 in of water by adding weights to the counterweight and then opening the prover valve and allowing the bell to drop under this pressure. If the bell does not drop, the obstruction should be located and removed. The latter test is made at a prover pressure of 1.5 in of water. When the bell is raised to the highest position, allow it to drain for 3 minutes. Then open the prover valve and drop the bell at a rate of 7.5 to 8.0 in per minute and observe the pressure variation on a gage capable of responding to dynamic pressure conditions and being read to 0.01 in of water pressure.