15th Flow Measurement Conference (FLOMEKO) October 13-15, 2010 Taipei, Taiwan

“Prover-by-prover”: CALIBRATION OF PROVERS USING

COMPACT PROVER AS MASTER METER

Valmir Ruiz 1, Marcos Tadeu Pereira 1, Nilson Massami Taira 1, José Fintelmam 2, , Danieli Guimarães 2
1
Instituto de Pesquisas Tecnológicas do Estado de São Paulo, São Paulo, Brazil, [email protected]
2
Hirsa - Sistemas de Automação e Controle Ltda, Rio de Janeiro, Brazil, [email protected]

Abstract: The article describes the calibration of petroleum products provers with a
compact prover as reference (master prover), using the same fluid in the same
conditions as used in the process line.

Keywords: Prover, Proving Systems, Calibration, Metering Systems

1. Introduction

The “Brazilian Technical Regulation for Measurement of Petroleum and Natural Gas” is
in use since June 2000, and companies that transfer fluids through ducts must have their metering
systems periodically calibrated according to the rules of this Technical Regulation. In the case of
transfer of large quantities of expensive fluids, the reference metering system in the petroleum
and petrochemical is usually a prover, either ball or compact, which is installed and aligned with
a totalizing flow meter, the custody transfer meter.

The prover calibration method described in the API – MPMS - Chapter 12.2 –Part 4 –
Calculation of Base Prover Volumes by the water Draw Method, is s the water draw, which
determines its volume either by using a standard test measure or by the gravimetric method
(weighing device), using water as calibrating fluid.

To use this method is necessary to disconnect the prover from the metering station,
normally isolating the prover from the system by means of a blind plate. The system must be then
be decontaminated, i.e., all the working fluid must be drained, and washed. After that, it is
necessary to assemble a bench composed by re-circulation pumps, a reservoir with enough
potable water to fulfill the test line and a standard test measure or a weighing tank. This requires
significant time and work, which implies that this kind of test is quite expensive.

To overcome these difficulties, an alternative calibration methodology was used, allowing

the calibration to be performed with the same fluid that flows through the piping, with the
advantage to save time and money in the process, generating at the end an acceptable global
uncertainty, within the limits of the regulations.

2. Methodology of Calibration
The calibration of provers using the "prover-by-prover" method consists basically in the
determination of the base prover volume using another calibrated prover as standard, and an
intermediate meter between both provers, to be calibrated by both provers and then extrapolating
the global uncertainty. This meter will be the responsible for the transference of the volume
values, as it is mentioned in the OIML R 119 - Pipe provers for testing measuring systems for
liquids other than water.

1
Copyright @ FLOMEKO 2010
15th Flow Measurement Conference (FLOMEKO) October 13-15, 2010 Taipei, Taiwan

Figure 1 – Calibration set-up lay-out, in the field.

T P
Master prover
Portable compact prover Optical
switches

Master
meter

Electronic meter
proving counter

Electronic meter
proving counter Flow reversal valve
with seal check facility Sphere
for master meter Pipe prover to be calibrated
T P D1

Vent valve
Detector Switch
supply line
T P D2

Return
line

In the tests conducted at the facilities of a petroleum refinery, a conventional ball prover with
3600 litres (prover under calibration) was calibrated with a compact prover of 94 litres (master
prover), using a 6 inches turbine meter as the transfer standard.

The mobile prover was assembled in series with the measurement station which contains the
prover to be calibrated, with all secondary instruments installed, composed of the following
items:
 Flowmeter, which was a turbine meter;
 Static pressure transmitter installed near the turbine meter;
 Temperature transmitter, sensor installed in the turbine meter.

These instruments were connected to the flow computer of the master prover (the compact
prover). These outputs are usually installed on the flow computer assembled in the measurement
station.

2
Copyright @ FLOMEKO 2010
15th Flow Measurement Conference (FLOMEKO) October 13-15, 2010 Taipei, Taiwan

Once the mobile prover is installed, the calibration system will function as a measurement
station composed of the two provers and a turbine meter.

The calibration of the turbine (master meter) installed on the measurement station should then
take place, using both provers in each run.

The meter calibration consists on the meter factor (MF) determination and its calculation is
presented below.

Combining the volume equations of the turbine meter and prover, the mathematical model
which is used for the determination of the MF is obtained. This model is obtained from both
provers, the master prover (mp) and the prover under calibration (puc). The index p refers to any
of the provers, the index m refers to the turbine meter.

Vp15
MF 
Vm15

Vp15  VBasep15  CTS p  CPS p  CTL p  CPL p

NP
Vm15   CTL m  CPL m
Kn

VBasep15  CTS p  CPS p  CTL p  CPL p

MF 
NP
 CTL m  CPL m
Kn

on which:

MF : meter factor of the turbine meter.

VBasep15: prover volume at 15ºC.
Vm15 : total volume indicated at the turbine meter, corrected to the reference condition of 15ºC.
Np : number of pulses generated by the turbine meter.
Kn : nominal k factor of the turbine meter (nº of pulsos/m³).
CTLm : correction factor that corrects the volume according to fluid thermal expansion or
contraction due to differences in operating and reference (15oC) temperatures in the
meter.
CPLm : correction factor that corrects the volume that flows by the meter due to fluid
compressibility in operating pressure, in relation to the reference pressure (zero).
CTSp : correction factor that corrects the prover volume according to thermal expansion or
contraction, due to differences in operating and reference (15oC) temperatures.
CPSp : correction factor that corrects the prover volume according to elastic deformation of the
materials of the prover, caused by the operating pressure of the prover.
CTLp : correction factor that corrects the volume due to the fluid thermal expansion or
contraction caused by the difference between operating and reference temperatures in
the prover chamber.
CPLp : correction factor which corrects the prover volume due to fluid compressibility or
expansion caused by the operating pressure inside the prover chamber when compared to
the reference pressure (zero).

3
Copyright @ FLOMEKO 2010
15th Flow Measurement Conference (FLOMEKO) October 13-15, 2010 Taipei, Taiwan

The “prover-by-prover” calibration method made the comparison between the results obtained
in the calibrations of the meter installed on the measurement station realized with both provers, in
the same operating conditions, and it is therefore important that the information contained on the
flow computers of both systems, be correctly configured with the same fluid specification, fluid
density, correction tables for the calculation of CTL and CPL, since these calculations are
implemented through the flow computers algorithms.

The turbine meter was calibrated realizing 5 runs for each flow rate. Two different flow rates
were used: operating flow rate and another at least 25% below this value.

The calibration of the turbine meter used as reference must take place simultaneously with
both provers installed in series. Once the meter has been calibrated and the calibration reports
from both flow computers obtained, with all corrections made, it is possible to calculate the new
volume for the prover that is being calibrated, using the volume of the master prover as reference.

3 Calculation of the new volume for the prover under calibration.

As explained, it was used a turbine meter to transfer the values of the master prover volume
(compact prover used as reference) to the volume of the prover under calibration.

The method should fulfill the following requirements:

The MF of the turbine meter obtained using the prover under calibration must equal the MF
obtained with the master prover used as reference.

MF15 Masterprover  MF15 New MFprover under calibration (I) and

MF15 Masterprover
 FC or
MF15 actual MFprover under calibratio n

MF15 New MF prover under calibration  Fc  MF15 actualMFproverunderc alibration (II)

And:

FC = Correction factor to new MF

MF15Master prover = Turbine meter factor obtained from the compact prover used as reference
corrected to the condition of 15°C.
MF15 actual MF prover under calibration = Turbine meter factor obtained from the actual volume (“actual” in
this case means the volume in the template of the prover) of the
prover under calibration corrected to the condition of 15°C.
MF15 New MF prover under calibration = Turbine meter factor obtained from the actual MF prover under
calibration corrected by the MF to the condition of 15°C.

Substituting (I) in (II)

MF15 New MF prover under calibration  MF15 actual MF prover under calibration  FC (III)

4
Copyright @ FLOMEKO 2010
15th Flow Measurement Conference (FLOMEKO) October 13-15, 2010 Taipei, Taiwan

V15 Newvolumep roverunder calibration

MF15 New MF prover under calibration  (IV)
V15 m
V15 actualvolu meproverundercalibration
MF15 actual MF prover under calibration  (V)
V15 m
Substituting (IV and V) in (III)

V15 Newvolumep roverundercalibration V15 actuavolumeproverund ercalibration

  FC
V15 m V15 m

Then:

V15 New volume prover under calibration  V15 actual volume prover under calibration  FC

So:

V15New volume prover under calibration = new volume for the Ball prover determined on the calibration.
V15actual volume prover under calibration = volume of the ball prover used in the calibration.

4. UNCERTAINTY BUDGET

The model for the calculation is:

VBasep 15  CTS p  CPS p  CTL p  CPL p

MF15 Masterprover 
NP
 CTLm  CPLm
Kn

MF15 Masterprover
 FC
MF15 actual MFprover under calibratio n

V15 New volume prover under calibration  V15 actual volume prover under calibratio n  FC

5
Copyright @ FLOMEKO 2010
15th Flow Measurement Conference (FLOMEKO) October 13-15, 2010 Taipei, Taiwan

UNCERTAINTY BUDGET - PROVER UNDER CALIBRATION

STANDARD
u (xi) DEGREES OF PROBABILITY SENSITIVITY CONTRIBUTION
UNCERTAINTY SOURCE, Xi SIMB. ESTIMATE, xi UNCERTAINTY ci . u[Xi] [ ci . u (Xi)]2
VALUE FREEDOM DISTRIBUTION COEFICIENT , ci (%)
u(xi)

Prover volume at 15ºC (Vbasep15), in liters Vbasep15 0,0945868 u(Vbasep15) 4,7E-06 infinity Normal 10,63224 5,03E-05 2,53E-09 20,43%

Corr. for the effect of temp. on steel in a prover (CTSp) CTSp 1,0007 u(CTSp) 6,80E-05 5,885E+08 Normal 1,00493 6,83E-05 4,67E-09 37,71%

Corr. for the effect of pressure on steel prover (CPSp) CPSp 1,0000 u(CPSp) 1,57E-06 1,282E+14 Normal 1,00562 1,58E-06 2,49E-12 0,02%

Corr. for compressibility of fluid in a prover (CTLp) CTLp 0,9830 u(CTLp) 4,51E-05 2260481 Normal 1,02301 4,62E-05 2,13E-09 17,22%

Corr. For compressibility of fluid in prover (CPLp) CPLp 1,0008 u(CPLp) 2,48E-06 1,066E+10 Normal 1,00489 2,49E-06 6,21E-12 0,05%

Pulses number (Np) Np 534,6404 u(Np) 0,1691 2 Normal 0,00018 2,99E-05 8,96E-10 7,24%

K factor, pulses per unit volume (Kn) Kn 5680,000 u(Kn) 0 infinity rectangular -0,00188 0,00E+00 0 0%

Corr. for the effect of temp. on a liquid in the meter (CTLm) CTLm 0,98305 u(CTLm) 4,51E-05 1,357E+06 Normal -1,02301 -4,62E-05 2,13E-09 17,23%

Corr. for compressibility of liquid in meter (CPLm) CPLm 1,0008 u(CPLm) 3,41E-06 450 Normal -1,00489 -3,43E-06 1,18E-11 0,10%

Meter factor MF 1,0057 381 Normal 0,00011 100%

Meter factor MF 1,0057 u(type B) 1,11E-04 381 Normal 1,00000 1,11E-04 0,000000012 9,58%
standard
MF 1,00532 3,42E-04 5,000E+00 Normal 1,00000 3,42E-04 0,000000117 90,42%
Average meter factor deviation(MF)

6 0,00036 100%

Coverage factor K = 2,52

Expanded uncertainty (MF) = 0,00090

Expanded uncertainty (%) = 0,09% Meter factor = 1,00532 ± 0,00090

6
Copyright @ FLOMEKO 2010
15th Flow Measurement Conference (FLOMEKO) October 13-15, 2010 Taipei, Taiwan

5. RESULTS

Three different ball provers were calibrated using the method, each one with a different fluid:
Diesel oil, propane and LPG-liquefied petroleum gas, using the same compact prover as master
meter.

The results presented refer to the temperature of 15C, and 101325 Pa. Thermal volumetric
expansion coefficient of the prover´s tube (mild carbon) Gc = 3,35E-05/oC. Modulus of elasticity
of the prover´s tube material (mild carbon steel) E = 206800000 kPa.

Table 1 – Results obtained in the calibration of the ball prover using Diesel oil

Run Flow Prover Volume Prover Volume Deviation from

No rate at 15°C original at 15°C new the original
volume
(m3/h) (Liters) (Liters) (%)
1 311 3603.42407 3613,57 0,28
2 309 3603.42407 3614,24 0,30
3 216 3603.42407 3613,99 0,29
Average 3613,93
Maximum deviation 0,019 %

The expanded uncertainty associated to the average volume is of 3,25 Liters or 0,09%
Note: The expanded uncertainty shown is based in a combined standard uncertainty multiplied
by a coverage factor k = 2,52, which gives an approximated 95% confidence level.

Table 2 – Results obtained in the calibration of the ball prover using propane

Run Flow Prover Volume Prover Volume Deviation from

No rate at 15°C original at 15°C new the original
volume
(m3/h) (Liters) (Liters) (%)
1 27 396,2086 395,79 -0,11
2 27 396,2086 395,79 -0,10
3 20 396,2086 395,79 -0,10
Average 395,79
Maximum deviation 0,001 %

The expanded uncertainty associated to the average volume is of 0,36 Liters or 0,09%

7
Copyright @ FLOMEKO 2010
15th Flow Measurement Conference (FLOMEKO) October 13-15, 2010 Taipei, Taiwan

Table 3 – Results obtained in the calibration of ball prover using LPG

Run Flow Prover´s Prover´s Deviation from

No rate Volume at 15°C Volume at 15°C the original
original new volume
(m3/h) (Liters) (Liters) (%)
1 247 1888,2 1888,40 0,01
2 244 1888,2 1888,37 0,01
3 224 1888,2 1888,41 0,01
4 179 1888,2 1888,89 0,04
5 179 1888,2 1888,88 0,04
Average 1888,59
Maximum deviation 0,028 %

The expanded uncertainty associated to the average volume is of 1,70 Liters or 0,09%

6. CONCLUSION
This paper shows the possibility of applying this calibration method, and that its uncertainty is
in accordance with the exactitude levels demanded by Brazilian law. It is, though, possible to
achieve better uncertainty levels as the methodology is used and improved.

REFERENCES

[1] API – MPMS - Chapter 12.2 – Calculation of Petroleum Quantities Using Dynamics
Measurement Methods and Volumetric Correction Factors. Part 3 – Proving Reports.
[2] API – MPMS - Chapter 12.2 – Calculation of Petroleum Quantities Using Dynamics
Measurement Methods and Volumetric Correction Factors. Part 4 – Calculation of Base
Prover Volumes by the water Draw Method.
[3] ISO 7278-2 Liquid hydrocarbons - Dynamic measurement - Proving systems for volumetric
meters - Part 2: Pipe provers
[4] Regulamento Técnico de Medição de Petróleo e Gás Natural, aprovado pela Portaria
Conjunta ANP/INMETRO Nº 1 de 19 de Junho de 2 000.
[5] OIML R 119 – Pipe provers for testing measuring systems for liquids other than water

8
Copyright @ FLOMEKO 2010