A pipe prover is a pipe or cylinder whose volume is used to verify and/or calibrate flow measuring systems for liquids. A process of calibration is accomplished by a displacer (usually a sphere or piston) passing through the pipe that actuates a pair of position detectors defining the calibrated section within a certain time. The known volume of this section (base volume) is used as a standard to determine the instrumental error of the measuring system which is connected to the prover. The measured volume is corrected for the errors due to the differences in temperature and pressure of the liquid from the reference condition. The travel time between the detectors is used to determine the flowrate.

The following are generally used as pipe provers. Other types of pipe provers can be used when their metrological characteristics comply with the requirements of this Document.

The calibration of the base volume of a prover shall be carried out such that the expanded uncertainty of the calibration is significantly smaller than the acceptable criteria which is referred to as the maximum permissible error (MPE) in type approval tests and verification tests. The treatment of the expanded uncertainty and the acceptance criteria should follow , the applicable OIML Recommendation, and/or an international standard. The expanded uncertainty shall be determined using an appropriate coverage factor ($k$k) which is usually equivalent to two ($k=2$k = 2).

The calibration methods summarised below are to be used as a guideline. A detailed description of these calibration methods is also given in . National regulations should specify that pipe provers are to be calibrated when they are installed, and that they should be re-calibrated periodically at an interval fixed by these regulations. Pipe provers should also be re-calibrated when they have been repaired or modified (e.g. change of detectors, recoating, dismantling and/or re-installation).

The base volume of a pipe prover shall be determined on the basis of the reference conditions specified in national/regional regulations.

There are three methods for calibrating a pipe prover: the water draw method, the master meter method, and the dimensional method.

A measuring system shall be tested using one of the liquids marked on the data plate (or in the instruction manual) of the measuring system, or with a liquid whose flow characteristics are within the ranges of those of the marked liquids. All appropriate safety and environmental measures concerning the handling of the liquids shall be observed.

It shall be ensured that the base volume of the pipe prover is matched to the resolution of the measuring system and is adequate to meet the requirements of . If the measuring system is equipped with a pulse generator which does not generate enough pulses for the base volume, a pulse-interpolation method may be used in compliance with .

Appropriate temperature-measuring devices shall be used to ensure accurate temperature correction of the liquid volume. These devices shall be mounted in suitable locations on the prover and the measuring system. It is recommended to use the temperature-measuring devices with an expanded uncertainty of $0.2\mathrm{°C}$0.2 "unitsml(degC)" or less. Calibration certificates shall be provided for the devices.

Appropriate pressure-measuring devices (e.g. pressure gauges) shall be used to ensure accurate pressure correction of the liquid volume. The devices shall be mounted in a suitable location on the prover and on the measuring system in order to measure the pressure of the liquid with the necessary accuracy. It is recommended to use the pressure-measuring devices with an expanded uncertainty of $0.05\mathrm{MPa}$0.05 "unitsml(MPa)" ($0.5\text{ bar}$0.5 " bar") or less. Calibration certificates shall be provided for the devices.

The electronic devices including detectors, pulse generator, and pulse counter shall be inspected to ensure correct operation.

A sufficient number of preliminary runs shall be carried out before the actual verification/calibration runs in order to ensure complete elimination of any gases, which may be contained in the prover and associated piping, and to ensure that the liquid temperature is stable and uniform. The measuring system shall be checked to ensure that leaks do not exist.

The total number and the range of flowrates applied for verification/calibration are specified in the applicable OIML Recommendation(s) for the measuring system.

The number of runs to be carried out at a particular flowrate is specified in the applicable OIML Recommendation(s) for the measuring system. In general, the number of runs is more than two in order to ensure the good repeatability of the measuring system as well as to verify whether each result meets the maximum permissible error (MPE).

The errors of the measuring system may be calculated either

1. from the meter factor determined according to 7, or

2. from the following equations in which the second order terms are neglected:

   ${\displaystyle E=E\prime +E_{\alpha }+E_{\beta }+E_{\gamma }+E_{\delta }}$E = E' + E_alpha + E_beta + E_gamma + E_delta ${\displaystyle E\prime =[(V_{\text{lm}}-V_{\text{B}})/V_{\text{B}}]\times 100}$E' = [(V_"lm" - V_"B") // V_"B"] xx 100 ${\displaystyle E_{\alpha }=\alpha \times (t_{\text{lp}}-t_{\text{lm}})\times 100}$E_alpha = alpha xx (t_"lp" - t_"lm") xx 100 ${\displaystyle E_{\beta }=\beta \times (t_{\text{s}}-t_{\text{lp}})\times 100}$E_beta = beta xx (t_"s" - t_"lp") xx 100 ${\displaystyle E_{\gamma }=\gamma \times (p_{\text{lm}}-p_{\text{lp}})\times 100}$E_gamma = gamma xx (p_"lm" - p_"lp") xx 100 ${\displaystyle E_{\delta }=-\delta \times p_{\text{lp}}\times 100}$E_delta = - delta xx p_"lp" xx 100

   $E$E

   is the instrumental error of the measuring system, in %

   $E\prime$E'

   is the uncorrected error, in %

   $E_{\alpha }$E_alpha

   is the temperature correction for the test liquid, in %

   $E_{\beta }$E_beta

   is the temperature correction for the pipe prover, in %

   $E_{\gamma }$E_gamma

   is the pressure correction for the test liquid, in %

   $E_{\delta }$E_delta

   is the pressure correction for the pipe prover, in %

   $V_{\text{lm}}$V_"lm"

   is the volume indicated by the meter, in $L$"unitsml(L)"

   $V_{\text{B}}$V_"B"

   is the base volume of the pipe prover, in $L$"unitsml(L)"

   $t_{\text{lp}}$t_"lp"

   is the liquid temperature in the pipe prover, in $\mathrm{°C}$"unitsml(degC)"

   $t_{\text{lm}}$t_"lm"

   is the liquid temperature in the meter, in $\mathrm{°C}$"unitsml(degC)"

   $t_{\text{s}}$t_"s"

   is the reference temperature of the pipe prover, in $\mathrm{°C}$"unitsml(degC)"

   $p_{\text{lp}}$p_"lp"

   is the gauge pressure of the liquid in the pipe prover, in $\mathrm{kPa}$"unitsml(kPa)"

   $p_{\text{lm}}$p_"lm"

   is the gauge pressure of the liquid in the meter, in $\mathrm{kPa}$"unitsml(kPa)"

   $\alpha$alpha

   is the cubic expansion coefficient of the test liquid due to temperature, in ${\mathrm{°C}}^{-1}$"unitsml(degC^-1)"

   $\beta$beta

   is the cubic expansion coefficient of the pipe prover due to temperature, in ${\mathrm{°C}}^{-1}$"unitsml(degC^-1)"

   $\gamma$gamma

   is the compression coefficient of the test liquid, in ${\mathrm{kPa}}^{-1}$"unitsml(kPa^-1)"

   $\delta$delta

   is the pressure expansion coefficient of pipe prover, in ${\mathrm{kPa}}^{-1}$"unitsml(kPa^-1)"

   The four coefficients may be determined as shown below:

   $\alpha$alpha

   Refer to or for petroleum products; refer to ISO 8222 for water

   $\beta$beta

   $33\times {10}^{-}6{\mathrm{°C}}^{-1}$33 xx 10^-6 "unitsml(degC^-1)" for mild steel, $51\times {10}^{-}6{\mathrm{°C}}^{-1}$51 xx 10^-6 "unitsml(degC^-1)" for stainless steel

   $\gamma$gamma

   Refer to the coefficient $C_{\text{pl}}$C_"pl" of 5.4

   $\delta$delta

   Refer to the coefficient $C_{\text{ps}}$C_"ps" of 5.3

   To obtain a more accurate result, the following equation can be used without neglecting the second order terms:

   ${\displaystyle E=(\frac{V_{\text{lm}}}{V_{\text{B}}}\cdot \frac{[1-\alpha (t_{\text{lm}}-t_{\text{lp}})][1+\gamma (p_{\text{lm}}-p_{\text{lp}})]}{[1+\beta (t_{\text{lp}}-t_{\text{s}})](1+{\delta }_{\text{lp}})}-1)\times 100}$E = ( (V_"lm")/(V_"B") * ([1-alpha(t_"lm"-t_"lp")][1+gamma(p_"lm"-p_"lp")])/([1+beta(t_"lp"-t_"s")](1+delta_"lp")) - 1 ) xx 100