Online Oil-in-Water Monitoring Experience - Enviro

Online Oil-in-Water Monitoring Experience
Arne Henriksen,
Principal Engineer,
Separation System and Subsea Technology, Research Centre Porsgrunn, Norway
2
Controlling and reporting oil discharge to sea, what is best?
BRAGE monitor
At-line sampling for Lab analysis
On-line sampling and analysis
3
Introduction
OSPAR CONVENTION FOR THE PROTECTION OF THE MARINE ENVIRONMENT OF THE NORTH-EAST ATLANTIC
MEETING OF THE OFFSHORE INDUSTRY COMMITTEE (OIC)
EDINBURGH: 14-18 MARCH 2005
Proposals for new methods of analysis and harmonised methods for sampling of oil in produced water
Presented by the United Kingdom
This document invites OIC to prepare an OSPAR Agreement on alternative method acceptance criteria
and harmonised methods for sampling of oil in produced water based on the attached report.
OIL IN PRODUCED WATER ANALYSIS
– ALTERNATIVE METHOD ACCEPTANCE AND GENERAL GUIDELINES ON SAMPLE TAKING AND HANDLING
Authors
Ming Yang, NEL
Steven Tulloch, OPUS
4
Implementing the ISO Mod Method using an online Monitor (OSPAR doc.)
[3.1] General Approach/Steps
1.
Select and procure an suitable online monitor
2.
Properly install the online monitor
3.
Select an onshore laboratory where the ISO Mod method will have been established and acceptable
to the regulators
4.
Establish an initial calibration / correlation over a period of time by taking replicate samples which
are analysed by both the online monitor and the ISO Mod method
5.
On-going validation of the correlation
6.
On-going routine calibration checks using a control chart
7.
Data for reporting
5
[1.] Select and procure an suitable online monitor
Optek, model TF16
Roxar, OiW monitor
Jorin, ViPA model MZ-3
ProAnalysis, Argus 100
Advanced Sensors,
OiW-EX100/1000 monitor
Teledyne Analytical instruments,
6650 OiW monitor
Sigrist Process-Photometer,
OilGuard EX
Turner Designs Hydrocarbon Instruments,
TD-4100 XD
6
[1.] Test Rig at StatoilHydro Research Centre in Porsgrunn
Water Test Rig:
Direct seawater input
Heater
Pump
Oil reservoir
Injection pumps
Choke
Test section
Valves
Drain with separator
Pressure measurements
Flow measurement
Temperature measurements
7
[1.] Online Monitors Test 2007 in Porsgrunn
8
[2.] Properly install the online monitor [4.2 Specific Guidelines for Sampling]
Guideline 1: Sample Point Location
The sample point for the collection of samples used for legislative monitoring, should conform to the following
criteria:
1. The sampling point shall be immediately downstream of the final oil-water separation, and from, or just downstream of,
a turbulent region.
2. The siting of the sample point should be approved by the regulator and should not be moved subsequently without
permission.
Guideline 2: Design of Sample Point
Online Sampling (with an Oil-in-Water monitor)
1. Where online systems require a side stream feed to the monitor, the feed should be provided by way of a dedicated
sample connection. A centre line pitot should be used to ensure that the sample is representative. Depending on the
available head and the monitor flow requirements, the sample may need to be pumped to the monitor.
2. Where online systems require the installation of a probe in to the main pipe, the location should be such that there
is confidence that the probe is exposed to representative flow conditions.
3. The distance between the sample point and any external online monitor, should be minimised as far as practically possible.
9
[2.] Sampling challenges for online OiW monitoring
10
[2.] Sampling challenge for online OiW monitoring
11
[2.] Online OiW monitor after the degassing tank at Brage platform
Light from the UV laser
in the sample chamber
Advanced Sensors OIW-EX1000 monitor
12
[2.] Reliable Automatic Cleaning System
Automatic cleaning system of the sapphire glass in front of the light probes based on ultrasonic wave
generated by the transducer. Result after 30 days continuing measurements.
13
[4.] Establish an initial calibration / correlation over a period of time by taking
replicate samples which are analysed by both the online monitor and the
ISO Mod method
Initial calibration / correlation:
In order to calibrate an online monitor against the ISO Mod method, a three-calibration points
approach is suggested. These three calibration points shall ideally correspond to high, mid-scale and low readings within
the monitor measurement range, which should at least cover 0 to 100 mg/l. As guidance, “low” means a reading in the
10-25% of the measurement range, “mid-scale” in the 40-60% and high in the 75-90% respectively. For each calibration
point, at least 3 parallel analyses / readings should be respectively taken by the online monitor and also by the ISO Mod
method.
Correlation should be obtained by plotting the average reading from the monitor at a time when samples are
withdrawn against the average reference value and then fit the points using linear regression. One must realise that
establishing this initial calibration/correlation may take weeks instead of days. Such a calibration may also include drifting
which may occur during the time period when this initial calibration is established.
During the initial calibration / correlation period, for the purpose of reporting daily samples should be taken and sent
onshore to be analysed. Alternatively another bench top analyser which has already been calibrated against the ISO Mod
method can be used.
14
[4.] Initial Calibration can be done in the Water Test Rig
Water Test Rig:
Direct seawater input
Heater (max 60 0C)
Pump
Oil reservoir
Injection pumps (oil, chemicals, solid)
Choke (oil droplet size, gas bubbles size)
Test section
Valves
Drain with separator
Pressure measurements (max 25 bar)
Flow measurement
Temperature measurements
15
[4.] Example of Calibration curve (mix of lab and offshore)
600
Constant
30.127671
Reg. Coeff.
8.702881
-0.040003
550
500
adv-09
adv-10
ME:
35.794364
adv-08
450
adv-07
400
adv-06
350
adv-05
24
300
Flu
adv-04
25
250
33
26
adv-03
200
29
150
31
32
100
30
28
23
adv-02
36
34
adv-01
27
37 38
35
50
0
null
-10
0
10
20
30
40
50
GC-ppm
BACCOS version 3.0 c:\brage_~1\brage_~1\baccos~1\calm02a
60
70
80
90
100
110
16
[4.] Sample cycles for Advanced Sensors OIW-EX1000 monitor
ppm
Flu
Responser
Cleaning – Homogenisation –Settling - Measuring [10 min interval]
12:44:00
12:47:02
12:50:04
12:53:06
12:56:08
12:59:10
Tid
BACCOS version 3.0 TMP-2
13:02:12
13:05:14
13:08:16
13:11:18
13:14:20
Measurements every second
04:56:36
04:51:27
04:46:18
22-25 40-41
04:41:09
04:36:00
04:30:51
04:25:42
04:20:33
04:15:24
04:10:15
04:05:06
03:59:57
03:54:48
03:49:39
03:44:30
03:39:21
03:34:12
03:29:03
03:23:54
03:18:45
03:13:36
03:08:27
38-33
03:03:18
02:58:09
02:53:00
02:47:51
02:42:42
02:37:33
02:32:24
02:27:15
02:22:06
02:16:57
33-35
02:11:48
02:06:39
32-39
02:01:30
01:56:21
01:51:12
34-33
01:46:03
01:40:54
01:35:45
01:30:36
01:25:27
01:20:18
01:15:09
01:10:00
response ppm
17
[4.] Results from spot samples, initial calibration / correlation
28. April, time 01:30 to 05:00
100
90
80
70
60
50
40
30
20
10
45-49
0
18
[5.] On-going validation of the correlation
[6.] On-going routine calibration checks using a control chart
Once an initial calibration/correlation is established, the online monitor’s performance should be continuously monitored
to confirm satisfactory operation and to detect any drift from the initial calibration.
It is suggested that routine calibration checks are carried out, initially once a week. Also validation should be done from
time to time to confirm that the initial calibration is still valid. Procedures involved in validation and routine calibration
checks are similar. Calibration curve / correlation shall consider to be valid if the difference between its readings
(after converting to the reference equivalent using the calibration curve / correlation) and the ISO Mod reference method
does not exceed the reproducibility of the ISO Mod method.
Routine calibration checks are carried out by analysing parallel samples as in establishing the initial calibration,
i.e. taking three samples during which readings from the online monitor are recorded. Then the difference in average
between the reference equivalent from the online monitor and the reference method is plotted on a control chat.
19
[5, 6] Validation results from Brage
•
Spot samples
Test period 12. March to 15. April
102 spot samples analysis: On-line method and GC method
60.0
– Mean deviation between GC method and
for 102 spot samples
Conc. OIL ppm
on-line method: - 4.2 mg/l (std dev. 5.0 mg/l)
50.0
40.0
LAB
30.0
ON-LINE
20.0
10.0
” Monthly reporting” 12. March – 15. April
11.27-00
05-47-30
23-13-15
18-21-30
11-26-00
06-07-40
23-49-50
18-13-38
11-41-30
05-47-30
23-46-00
16-45-44
23-34-00
23-22-00
17-23-00
11-06-00
06-16-00
23-30-00
06-05-00
23-59-00
17-42-00
0.0
– Spot samples and same time period on-line
Sample time
– GC method (average of 4 spot samples/day):
Scatte r plot, GC m e thod and On-line m e thod
13.5 mg/l
60
50
– On-line method (average of 4 spot sampling
time/day):
17.7 mg/l
On-line Analysis (ppm)
•
40
30
20
10
– On-line analysis (mean value over 24 hours):
0
17.8 mg/l
0
5
10
15
20
GC Analys is (ppm )
25
30
35
40
20
[7.] Data for reporting
Online monitors gather data continuously. However for regulatory compliance monitoring, only two grab samples per day
currently require to be analysed. Therefore in theory not all the information generated by the on-line monitor needs
to be reported. There are a number of possible ways which may be considered for reporting:
(i)
select two specific time of the day, and take the two oil-in-water results closest to that time for reporting;
(ii)
select two specific short periods of time (say 5 minutes or a time duration which allows to take 3 consecutive
measurements) each day and then take the average of the online monitor readings for reporting;
(iii)
average all the results obtained daily and then report the average result.
Of the three options, option (ii) is probably better than the other two as it will smooth out the readings and at the same
time allow plenty of time for the instrument to be maintained daily if required.
21
[7.] Data for reporting
Further checks on the equivalence of the results obtained from the online monitor and the reference method using statistical
significance tests (F tests and Student-t tests) can be carried out.
Dealing with online monitor break-down: if an online monitor is used for compliance monitoring / reporting, it is suggested
that at least a laboratory bench top analyser which has been calibrated against the ISO Mod method should be available
as a back-up.
Responsibility: for online monitors, it is thought that it is better that the operator using the monitor takes the full responsibility
in establishing and validating the correlation between the alternative method and the ISO Mod method, and also the running
of the control chart.
22
Results 20-30 April
110
20
100
21
22
23
24
25
26
27
28
29
30
90
80
o il p p m
70
60
50
40
30
20
10
0
1
63 125 187 249 311 373 435 497 559 621 683 745 807 869 931 993 1055 1117 1179 1241 1303 1365 1427 1489
on-line analysis (10 minutes intervals)
23
Results 1-8 May
100
1
2
3
4
5
6
7
8
90
80
Oil ppm
70
60
50
40
30
20
10
0
1
51
101
151
201
251
301
351
401
451
501
551
601
651
701
On-line analysis (10 minutes intervals)
751
801
851
901
951
1001
24
24 hours report (automatic)
25
Online oil in water measurements for real time process control
Monitor Screen at BRAGE Control Room
26
Online OiW monitoring at TROLL C platform (2007)
27
Online OiW monitoring at TROLL C platform (2007)
28
TROLL C trend measurements for 35 days after the start up (July – August)
•
The installation was a test application June 2007
•
OiW monitor measuring the inlet water stream to the
600
500
degassing tank
400
•
The calibration is not optimised (0 to 600 ppm)
300
•
The platform used the 4-20 mA signal (ppm) today for
process responding when the oil concentration
200
rapidly increase
100
Need a better calibration (in progress)
•
Instrument has been running for 9 months
without any maintenance
0
05
:2
05 2
:2
05 8
:5
05 2
:5
05 5
:5
05 0
:0
22 5
:3
06 0
:0
06 8
:0
06 8
:0
06 0
:0
04 0
:0
05 2
:2
03 6
:5
03 8
:5
03 0
:59
•
Spot_GC metoden
OiV analysator
Differansen
29
Production chemical in produced water (Full Scan Spectrometer)
500
si4503-10
si4503-50
si4503-100
2200
25ppmOIL
25ppmOIL+20ppmSI
Std. dev.
2000
450
Scale inhibitor
1800
400
1600
350
1400
300
1200
Intensity
250
In t
1000
200
800
150
600
100
400
50
200
0
0
370
410
449
489
528
568
607
647
686
726
wavelength, nm
BACCOS version 3.0 f:\advanc~1\databe~1\oilfie~1\chemmix0
765
805
844
884
923
963 1003
370 419 468 517 565 614 663 712 761 810 858 907 956 1005
wavelenghts nm
BACCOS version 3.0 f:\advanc~1\databe~1\oilfie~1\stat2
30
Conclusion: Online Oil-in-Water Monitoring Advantages
ENVIRONMENT (real-time analysis)
PRODUCTION (real-time analysis)
•
Oil discharge analysis for 24 hours
•
Better process control
•
Operator can handle quickly to process
•
Process optimization
disturbance and minimise oil to sea
for achieving
discharge
- higher oil production
•
reporting
•
- without increased oil discharges
Platform can use the analysis for SFT
Reduction in Lab work at the platform
•
Potential for analysis of
- higher oil concentration
- production chemicals
- Subsea application