Sie sind auf Seite 1von 31

Jon Moreau is Global Business Development Director for Cameron Measurement

Systems. He has been with the company for nearly 20 years having joined Jiskoot in
1995 prior to its acquisition by Cameron in 2008.

Jon has worked extensively with quality measurement and control systems and has
worked on and led a number of major projects leveraging sampling technology to
deliver a rapid return on investment in the Middle East, Asia, Europe and the Gulf of
Mexico.

Jon is a previous presenter at Measurement conferences in the Middle East and


Internationally. He is an active author of technical documents, many of which have
been published internationally and is also a contributor to working committees for
JON MOREAU quality measurement and control issues including sampling and blending.
Global Business
Development Director, Jon was instrumental in working with the technology team at Jiskoot to create many
Cameron of the return on investment models that are now extensively used for quality
measurement and blending.
Measurement
Systems
Proving
Proving

?
How do you manage risk?
Paper Abstract
The accuracy of a liquid hydrocarbon (crude oil, condensate etc.)
sampling system is a vital component of the performance of a
metering system and the ability to account for crude oil custody
transfer or allocation.

However, unlike the performance of flow meters which are


validated by proving, the accuracy of a sampling system is often
unknown, or not validated and is therefore not considered when
calculating the measurement uncertainty calculation of the overall
metering system performance. This can have a significant impact
on the revenues of an organisation of the size of KOC.

This papers reviews the methods for validating the performance


of a crude oil or condensate sampling system, considers the effect
of the performance of a sampling system on overall measurement
uncertainty. It includes a relative evaluation of the performance of
different types of sampling systems and considers the process
variables that have the most effect on sampling system accuracy
using a large data set of field performance proving results of
installed sampling systems
Potential quality measurement losses for Kuwait

$250

$200

$150
Loss pa (Million $)

$100

$50

$0
%0.025 %0.050 %0.100 %0.200 %0.500
Water content?
QUANTITY MEASUREMENT QUALITY MEASUREMENT
I.E. METERING I.E. SAMPLING

VOLUME LIQUID VOLUME OIL


+/- X% +/- Y%

METERING SAMPLING
UNCERTAINTY = X UNCERTAINTY = Y

WHEN ASSESSING THE VALUE OF QUANTITY AND QUALITY SOMETIMES YOU JUST NEED BOTH
What is the RATIO between the volume analyser in the lab for BS&W and a VLCC cargo?
A. <5,000,000:1
B. <10,000,000:1
C. >10,000,000:1

Crude oil cargo - 300,000 m3


10,000 = 1cc sample grabs taken per batch

300,000,000,000:1 10 litre sample taken to lab for analysis


1 100cc analysed for water - depending on method
So how do you Prove a sampling system

API and ISO define a water


injection test to prove the
uncertainty of the entire
sampling & analysis chain

Proves that the sampling


system operates within an
acceptable uncertainty
So how do you Prove a sampling system

Check results using analyzed


sample
Wlab = Winj + Wbaseline

Two sequential tests for


repeatability
Tolerance per standards
ISO 3171 /IP 6.2
API 8.2
Proving acceptance criteria API Two tests

Total Water Using Tank Gauges Using Meters


0.5 0.13 0.09
1.0 0.15 0.11
1.5 0.16 0.12
2.0 0.17 0.13
2.5 0.18 0.14
3.0 0.19 0.15
3.5 0.20 0.16
4.0 0.21 0.17
4.5 0.22 0.18
5.0 0.23 0.19

Proving to API standard


Only complies with API not ISO or IP!
Proving acceptance criteria ISO/IP

Proving to IP or ISO standard


Complies with API, ISO and IP (2 tests)
Proving to the API complies only with API
Sampling system certification test Test conditions

Minimum Norm Maximum


Flow Rate (m3/hr) 500 2500 5000
Viscosity (Cst @ process) 2.5 4.7 12
Density (kg/m3) 823 855 890

A B C

?
Sampling system certification test Test conditions

Minimum Norm Maximum


Flow Rate (m3/hr) 500 2500 5000
Viscosity (Cst @ process) 2.5 4.7 12
Density (kg/m3) 823 855 890

A
WHERE POSSIBLE WATER INJECTION PROVING SHOULD BE CARRIED OUT
AT THE WORSE DESIGN CONDITIONS FOR SAMPLING OF MINIMUM
FLOW RATE, DENSITY AND VISCOSITY
Sampling system certification test Test conditions

Test the system using most stringent conditions:


Lowest flow rate
Lightest viscosity / density

Things To Do First
Conduct the sample receiver mixing and handling component test
Conduct the sampling device component test

System Test Failure!


Cause of system certification failure can be any number of things
Eliminate as many component variables as possible beforehand

USE THE EXPERT SERVICES WHERE POSSIBLE


Typical Sampling System Proving Certificate
Typical Sampling System Proving Certificate Independently witnessed
Typical Sampling System Proving Certificate - America
Water injection tests provide the ability to compare the performance of different types of
sampling systems.

How do different types of sampling systems perform?


Inline Samplers
Fast Loop Samplers
Either combined with
Power Mixers (CoJetix)

Historic sampling system


certification data helps
to answer that question..
In-line sampling system test data
In-line sampling system test data all types of mixing
Typical Inline Sampling Systems Error in Relative Water Cut
0.50%
0.40%
0.30%
0.20%
Relative WC error %

0.10%
0.00%
-0.10%
-0.20%
-0.30%
-0.40%
-0.50%

Other Mean Error 95% confidence level

Mean error of about -0.1% relative water fraction. The scatter about this mean however is relatively large being about
0.4% meaning any one system result can have a relatively large variation from this mean.
In-line sampling system pre 1998 JetMix test data
Pre 1988 Inline JetMix Sampling Error in Relative Water Cut
0.50%

0.40%

0.30%

0.20%
Relative WC error %

0.10%

0.00%

-0.10%

-0.20%
-0.30%

-0.40%

-0.50%
JetMix pre 1988 Mean Error

The JetMix and Inline Sampling System combination was developed during the early 1980s, and during the early period
these systems went through significant JetMix Nozzle development that culminated in new Nozzle designs. Note the
scatter and the mean.
In-line sampling system post 1998 JetMix test data
Post 1988 Inline JetMix Sampling Error in Relative Water Cut
0.50%

0.40%

0.30%

0.20%
Relative WC error %

0.10%

0.00%

-0.10%

-0.20%
-0.30%

-0.40%

-0.50%
JetMix post 1988 Mean Error 95% confidence level
The combination of an active mixer and Inline Sampling System significantly improves the results from Inline Sampling
Systems. The results are comparable to that seen in the Fast Loop Sampling Systems results (covered in the next section)
Fast Loop (bypass) sampling system test data

ByScoop take-off
with integral
return
Fast loop (bypass) sampling system test data
Fastloop Sampling System Error in Relative Water Cut
0.50%
0.40%
0.30%
0.20%
Relative WC error %

0.10%
0.00%
-0.10%
-0.20%
-0.30%
-0.40%
-0.50%

Fastloop Mean Error 95% confidence level

The Fast Loop Sampling System has a marked increase in accuracy over Inline Sampling Systems. Reduces the mean
error to that of about -0.05% relative water fraction. Reduces scatter in the individual results to that of about 0.15%.
CoJetix (combined fast loop and JetMix) sampling system test data

JetMix
CoJetix (combined fast loop and JetMix) sampling system test data
CoJetix Sampling System Error in Relative Water Cut
0.50%
0.40%
0.30%
0.20%
Relative WC error %

0.10%
0.00%
-0.10%
-0.20%
-0.30%
-0.40%
-0.50%

CoJetix Mean Error 95% confidence level

Mean relative error of <-0.01% / Scatter less than 0.1%.


The CoJetix Sampling System design is the most accurate despite being installed in the most challenging locations.
Sampling systems comparison
Combined JetMix and CoJetix Error in Relative Water Cut
0.50%
0.40%
0.30%
0.20%
Relative WC error %

0.10%
0.00%
-0.10%
-0.20%
-0.30%
-0.40%
-0.50%
JetMix before 1988 JetMix after 1988
CoJetix Mean Error
95% confidence level
CoJetix sampling technology yields higher accuracies, narrower 95% confidence interval, and average results closer to 0%
relative error. Despite ISO 3171 Annex A mixing calculations showing that there is equal mixing using an Inline Sampling
System with passive mixing.
.
Summary
Sampling accuracy has a large effect on net oil measurement and unless you know
your sampling system performance losses can be significant
Not all sampling system designs have the same performance
ISO & API have a method for proving sampling system performance that is
international recognised and auditable
Market conditions are driving buyers and sellers to invest in reducing measurement
uncertainty
This is driving increasing challenges and changes in crude oil contract terms

Do you know the accuracy of your sampling system?


Thank you

Das könnte Ihnen auch gefallen