This document supports an Energy Institute Standard.

It is a record of the round robin study from which the precision of the test method was derived. It is available on request for study, however if the information in this document is intended to be used as part of a new document or presentation, the approval of the Test Method Standardization Committee Secretariat of the Energy Institute is required.

Marine Residual & Distillate Fuels Research Report IP 570 (Procedure A)

Determination of Hydrogen Sulfide in Fuel Oils Rapid Liquid Phase Extraction Method
SC-G-5

Author: Ian Mylrea, Chair EI SC-G-5 H2S Task Group

Dated 6th March 2012

Round Robin Study Report Summary A round robin was held using eight H2S Analyser Instruments and twelve samples in duplicate, to produce the repeatability and reproducibility of the test method. The samples used were distillate and residual fuel oils and components. Values of measured H2S concentration varied from 0 to 12.5 mg/kg. The resulting calculated expressions of repeatability and reproducibility (valid from 0.60 to 12.5 mg/kg) were as follows: Range of Results (mg/kg) 0 12.5 Repeatability (mg/kg) 0.3329 x0.55 (65) Reproducibility (mg/kg) 0.4459 x0.55 (95)

The round robin results have shown that IP 570 is suitable for measuring very low levels of H2S down to 0.6 mg/kg in the liquid phase. It is therefore recommended that the method be balloted with the inclusion of this new precision for approval with this Research Report. Introduction Energy Institute Test Method IP 570 was developed in response to demand from industry for a quick and reliable test to determine the level of hydrogen sulfide present in the liquid phase in fuel oils. IP 570 was first published in 2009 following an Energy Institute Round Robin in March 2009. Distillate fuels were omitted from the round robin due to there not being any suitable test specimens with H2S present. Although it is unusual to get H2S in distillate fuels it is not unprecedented and occurrences have been reported. In 2011 a distillate round robin was successfully completed and IP 570 was updated to include residual and distillate marine fuels. Also in 2011 Note 5 was added to the method regarding interferences, which can cause elevated readings. Further development of the method was required to remove interferences. This round robin included a new procedure and equipment to remove interferences. The information contained herein represents the finding of an Energy Institute round robin held at the Stanhope-Seta Laboratory, Chertsey, UK, on the 14th and 15th February, 2012. The trial compared the performance of 8 instruments testing 12 samples in duplicate and in a random order. In total 8 operators took part and a number of independent observers were present. The purpose of the round robin was to update the current test method IP 570 Determination of hydrogen sulfide in fuel oils Rapid liquid phase extraction method to include precision for the new procedure and equipment using the new vapour phase process outlined. Objectives The objective of the study was to generate repeatability and reproducibility data for the test method, to give an estimate of the performance of the instrument under typical laboratory conditions for Marine Fuels as per ISO 8217. Test Method The test method used for this study was titled IP 570/12a draft Determination of hydrogen sulfide in fuel oils Rapid liquid phase extraction method and is attached as Appendix B.

List of Participating Laboratories and Operators Owing to the unstable nature of H2S in marine fuels, all testing was done at a single site in two days. The laboratory used in this study was at Stanhope Seta , London Street, Chertsey, Surrey, KT16 8AP, UK. . The following people took part in the round robin. Name Anne Christine Barbier David Browbank Gary Ogborn Anna Prendergast. Willie Schermer Bert Decelle Simon Crozier Phil Mortimore Company Total (France) Lintec ITS (West Thurrock) ITS (West Thurrock) Shell ExxonMobil Nalco ExxonMobil

The following people were independent observers at the round robin. Comments received can be found in Appendix D. Name Victoria Dodd Jim Barker Alister Jackson Dav Basra Barbara Heyberger Company UK Energy Institute Innospec (TMS SC-G Chair) ExxonMobil ConocoPhillips Total (France)

Test Specimens Reference SA0718 SA0722 SA0701 SA0697 SA0711 SA0717 SA0704 SA0694 SA0707 Distillate or Residual Distillate Residual Residual Residual Residual Residual Residual Residual Distillate Type Gas Oil Decanted Oil RMG 700 Residual Fuel RMG 380 RMG 380 Residue from cracker RMG 380 F76 marine distillate Def Stan 91-4 RMG 500 F76 marine distillate Def Stan 91-4 DMA (ISO 8217) Viscosity (cSt) 3.25 (40oC) 83 to 120* (50 oC) 600 to 700* (50 oC) Not known (>300) 380 max* (50 oC) 335.4 (50 oC) 75 to 85* (50 oC) 380 max* (50 oC) 1.7 to 4.3* (40 oC) 500 max* (50 oC) 1.7 to 4.3* (40 oC) 3.5* (40 oC) Container ID 1 2 3 4 5 6 7 8 9 Day 2 AM Test Session

Day 1 AM

Day 1 PM

SA0709 SA0736

Residual Distillate

10 11 Day 2 PM 12

SA0439

Distillate

TABLE 1 ROUND ROBIN SAMPLES

*viscosity obtained from supplied information or grade specification During the round robin the Container ID was used as the identifier for the sample. Each container was provided with a label which showed the Container ID. Repeat tests were taken from the same container and in a random order.

Table 1 above describes the samples used in the round robin. All the samples except sample 9 and 11 were received directly from refiners. Sample 9 was sampled from a vessel. Sample 11 was purged with Hydrogen Sulfide. In order to minimise the possibility of the concentration of H2S altering significantly over the test period, the samples were divided up into four groups to ensure that the duplicate test result was obtained within a maximum of 4 hours of the first test result. Over two days there were four groups. Group 1 consisting of samples 1,2 and 3 were tested on the morning of Day 1. The duplicate (repeat) was also in a random order and did not necessarily follow the first sample. Group 2 consisting of samples 4,5 and 6 were tested on the afternoon of Day 1. Group 3 consisting of samples 7,8,9 and 10 were tested on the morning of Day 2. Group 4 consisting of samples 11 and 12 were tested on the afternoon of Day 2. No samples were tested on more than one day and no samples were tested in the morning and afternoon. Round Robin Protocol The protocol was approved at the Task Group meeting January 13th 2012. The table of samples was updated to include the actual samples used on the day. See Appendix A. Apparatus Eight H2S Analyser Instruments were used for the round robin testing. Each analyser also included a Vapour Phase Processor (VPP) as a permanent feature. As part of the instrument apparatus, 10ml disposable syringes with a short length of 6mm nylon tubing and glass test chambers were supplied. Syringes were used for taking the test portion . The size of the test portion taken was 2ml or 5ml depending on the H2S concentration. VPP Cartridges (sorbent specification SA4017/A), lot number 0001447701 were supplied. One cartridge was used for each test. The manufacturer quality controls the cartridges. A new syringe was used to draw the sample for each test. Software The H2S Analysers all utilised software versions 5.00/2.00. The VPPs all utilised software version 1.11.

Data Table 2 below is the data generated by each of the instruments used in the trial.

TABLE 2 ROUND ROBIN RESULTS

Repeats 1 and 2 are the first and repeat results. All results are in mg/kg. The results recorded as -999.9 were removed for the reasons stated below Instrument All 1022111 1022109 1022109 1022111 1009088 1009088 1022107 Sample Sample 2 Sample 7, Repeat 1 Sample 4, Repeat 2 Sample 1, Repeat 1 Sample 4, Repeat 1 Sample 4, Repeat 1 Sample 4, Repeat 2 Sample 11, Repeat 1 Reason Incorrect sample portion Incorrect sample portion Operator reported long delay recorded between adding sample and starting test Abnormal flow rate Operator reported unstable background/baseline and data supports Abnormal flow rate Abnormal flow rate Operator reported bubbling the sample when setting the test up

LIST 1 REMOVED RESULTS

The test methodology requires the user to predict the H2S level prior to test and then choose the appropriate volume of sample. As the samples were pre-tested advice was given on the volume of sample to use in the protocol. In the case of Sample 2, which had an expected level of H2S of just over 10 mg/kg it was advised to use a 2 ml test portion. However most results were < 10 mg/kg and hence the wrong test portion volume was used. Sample 2 was therefore eliminated. . Under normal laboratory conditions the sample test would have been repeated immediately, however in these cases this was not done. Future protocols shall include advice regarding this aspect. A long delay in starting the test was recorded against one of the test results, whereby the sample had been kept at an elevated temperature for a long period. The test method does not currently include any advice, but it is good practice to introduce the test portion into the apparatus just before commencing the test procedure to save degradation of the sample. It is suggested that an advisory note be added to the test method once the sample is in the test chamber start the measurement without delay to avoid possible sample degradation. This could be added under 11.4.8. The graphical outputs of three samples were checked in order to understand why the result was lower than expected. Upon inspection it was obvious that the graph was not shaped correctly and typical of a change in flow rate during the test or for the entire test. It is suggested that a typical graphical output is included in an appendix at the back of the test method to assist users. It is not fully understood why this occurred and there is an investigation taking place by the manufacturer into possible causes. During one test the operator had reported that the apparatus baseline/background was observed as changing from normal. Inspection of the data confirmed this. It is not fully understood why this occurred and there is an investigation taking place by the manufacturer into possible causes. On one sample the operator reported bubbling/disturbing when injecting the sample into the test vessel. This means that air in the syringe passed through the sample while it was added to the test vessel. This practice results in a loss of H2S. See 11.4.4 in the test method do not expel air from the syringe or pipette into the sample. A re-test was not done by the operator.

Statistical Evaluation on the Data The data was analysed according to ISO 4259 (IP 367) using the ASTM D2PP software package to generate values for repeatability and reproducibility. This analysis highlighted that, as previously found in IP570 precision studies, there was evidence of excessive variability according to the concept of Limit of Quantitation. This implies that differentiation between samples may be difficult at low levels of result (sample means below 0.6 mg/kg) The data in Table 2 was analysed according to ISO 4259 (IP 367) using the D2PP software package to generate values for repeatability and reproducibility. Full details of data used and the D2PP report can be found in Appendix C. The repeatability and reproducibility were found to be: Range of Results (mg/kg) 0 12.5 Repeatability (mg/kg) 0.3329 x0.55 (65) Reproducibility (mg/kg) 0.4459 x0.55 (95)

These relationships are summarised in Table 3 and Figure 1:

! ! " #! ! !# ! ! # ! # ! $"

! !
TABLE 3

& '

Thanks and acknowledgements Members of the Working Group and to the participants and observers of the round robin Royal Navy, ExxonMobil, Intertek, Total, ConocoPhillips, Chevron, Valero and Ineos for supplying the test samples Particular thanks to Stanhope Seta for the very generous use of their facilities

Appendix A

This document supports an Energy Institute Standard. It shall not be reproduced or circulated or quoted, in whole or in part, outside of the activities of the Standardization (SC) Committee of the Energy Institute except with the approval of the chairperson of the SC Committee.

ROUND ROBIN STUDY IP 570

Determination of hydrogen sulfide in fuel oils Rapid liquid phase extraction method 14th & 15th February, 2012

SC-G-5 Task Group on Measurement of H2S in the Liquid Phase

10

Ian Mylrea Task Group Chair Stanhope-Seta

ISSUE 1.5

Dear EI TMS G-5 Panel H2S Task Group member, As I am sure you aware that all our EI Panel activity in the EI TMS committees and panels is carried out on the understanding that the meetings discussions and shared documents are for Panel use only and are not to be shared outside that EI Panel membership, with the exception that within your organisation technical additional input can be shared/ supplied to enhance the method development.

11

Please observe this protocol of maintaining the activity/discussions involving EI Interlaboratory Studies (i.e. Robustness Studies & Round Robins), any data/observations derived from these studies are also deemed to be only for discussion within the EI Panel group activity (our Task Group), until the final agreement and release of any Official EI TMS Panel reports that have been approved by the EI Panel Membership, EI Subcommittee Chair and EI TMS for public release and discussion.

12

1 SUMMARY

The objective of this round robin study is to determine the precision of the test method in accordance with ISO4259.

The successful outcome will result in the current test method IP 570 Determination of hydrogen sulphide in fuel oils Rapid liquid phase extraction method being balloted to add precision for distillate and residual fuels using the new Vapour phase process methodology outlined.

The study will be carried out at the SA Laboratory in Chertsey, on 14th & 15th February 2012 using 8 production instruments and the same number of operators plus observers nominated by the Task Group members. 8 to 12 samples will be tested in duplicate and the EI statistician will determine the repeatability and reproducibility.

2 INTRODUCTION

Energy Institute Test Method, IP 570, is already used by industry to measure residual marine fuels, for which precision data was obtained in 2009. A second round robin in 2011 added precision for marine distillate fuels. The latest published version of the method is IP 570:2012. In October 2010, at the H2S Task Group Workshop, held at the Lloyds registry, information concerning interferences was submitted. Following studies by the manufacturer, a note (Note 5) was added to the test method regarding the possibility of interfering chemicals such as low boiling point thiols and alkyl sulfides elevating the readings. Research followed and a new process was developed and tested with a wide range of samples and interferences. The new process can be fitted, as an accessory, to the existing analyser.

ISO 8217:2010 requires the liquid phase Hydrogen Sulphide level to be determined according to this method. There is a phased entry for this part of the specification to enable industry to prepare itself for the limit of 2 mg/kg. The limit comes into force in July 2012.

8 production instruments, fitted with the vapour phase process accessory are being used for this round robin study.

13

The test requires a weighed 1 ml to 5 ml sample, depending on H2S concentration, and tests are completed in 15 to 25 minutes.

The development would not have been possible without the encouragement and support of Lloyd's Register Group FOBAS Service, Intertek OCA and major refiners.

The SC-G-5 Task Group on Measurement of H2S in the Liquid Phase met at the EI and agreed on the draft test method being used for this study.

3 TEST METHOD (attached)

Determination of hydrogen sulfide in fuel oils Rapid liquid phase extraction method

14

4 SAMPLES

There are 12 samples as detailed below. Samples were typically supplied in 5 litre epoxy lined containers, approximately 95% full. Distillate samples were supplied in 1 litre containers.

Reference SA0718 SA0722 SA0701 SA0697 SA0711 SA0717 SA0704 SA0694 SA0707 SA0709 SA0736 SA0439

Sample ID 1 2 3 4 5 6 7 8 9 10 11 12

Type Distillate Decant Oil Residual Fuel Residual Fuel Residual Fuel Residual Fuel Bottoms Residual Fuel F76 marine distillate Residual Fuel F76 marine distillate DMA

Nominal H2S (mg/kg) 3 to 6 10 to 12 0 to 3 0 to 5 0 to 3 0 to 3 5 to 8 0 to 3 0 to 3 0 to 3 10 to 12 0 to 2

Test volume req. (5ml) 5 2 (Two) 5 5 5 5 5 5 5 5 2 (Two) 5

The above table outlines the desired range of H2S levels for this round robin. This is for information only, as the actual H2S concentrations will depend on the samples available for the testing on the day of the study.

5 PROTOCOLS INSTRUCTIONS FOR PARTICIPANTS

5.1 Each operator will be designated an individual apparatus and printer, and be given a copy of the test method and this protocol.

15

5.2 Operators will be given training before the tests commence.

5.3 Due to the volatility of the H2S and the need to test the potential of the apparatus and not the sampling, samples are generally provided in 5L epoxy lined containers (except Sample 9, 10 and 11, which are supplied in 1 litre containers). For each sample there are 2 containers (except Sample 10 & 11). 1 container enabled pre-testing of the sample to ensure the desired range of H2S in the samples to be covered. The second container will be used by the participants for the first and duplicate test. Samples are supplied with minimum headspace from major refiners.

5.4 The sequence of testing will be in a random order and different for each operator.

5.5 Each operator will be given a results sheet that gives the required sequence of testing.

5.6 Testing should be in accordance with the test method, with these following points being of particular importance: 5.6.1 In order to minimise loss of H2S, do not homogenize or shake the sample before taking a test portion. 5.6.2 Once a sample has been taken, immediately tightly replace the container seal to prevent the release of further H2S. 5.6.3 The correct size of sample can be determined by reference to table 1 below: Table 1: Appropriate Test Volume for Expected H2S Concentration Expected H2S Concentration 0 - 10 mg/kg 10 - 20 mg/kg 20 mg/kg + Required Test Volume 5 ml 2 ml 1 ml

5.6.4 Ensure that the Input sample weight in grams screen is displayed before removing the sample vessel cap and introducing the test portion. 5.6.5 When introducing the sample into the test vessel, it is recommended that the operator lift the test vessel proud of the surface. This enables the level of the liquid to be seen which helps prevent the tip of the pipette or syringe touching the diluent oil.

16

5.6.6 Pipette or syringe the test portion into the test vessel ensuring that the pipette or syringe is held vertically to avoid sample adhering to the test vessel walls. 5.6.7 Once the sample has been entered into the equipment or vessel glassware, immediately refit the glassware screw cap tightly to prevent leakage.

17

Summary Guide to follow and run samples

1. Initiate cooling on the vapour phase processor (VPP) 2. Insert a new cartridge into the VPP and replace the lid 3. Allow 10 minutes before starting the test (Step 10 onwards) 4. Using a 20ml syringe put 20ml diluent oil into a clean test vessel 5. Place the vessel in the apparatus 6. Set the mode of the apparatus to enable bubbling 7. Wait for 5 minutes to allow diluent to warm up before obtaining the sample 8. Type in the Sample ID 9. Type in the Operator ID 10. Go to the Samples. a. choose the correct sample can from your test sheet. b. Obtain the sample using a clean 10ml syringe c. Draw the sample into the syringe slowly and carefully to avoid exposing the sample to a strong vacuum. d. Take 10 to 20 seconds to draw the sample in for example. Draw the sample beyond the 5ml mark. e. Make sure the syringe is at least 3cm below the liquid surface of the sample container f. Avoid drawing air through the sample during sampling. g. Wipe the syringe with paper towel h. Add a tube to the syringe to facilitate correct application of the sample later on (it may be convenient to add the tube before drawing the sample) 11. Go to the weighing balance a. weigh the syringe and tube and write down the Time in the Time column and the weight in the Full Syringe Weight column 12. Go to the apparatus. a. select the mode of the apparatus to enter the weight b. enter the weight of the sample syringe. c. check the weight is correct. d. press the enter button. e. open the red cap of the sample vessel and inject the sample f. replace the cap immediately 13. Go to the SAME weighing balance a. weigh the syringe and tube and write down the weight in the Empty Syringe Weight column 14. Go to the apparatus a. Enter the empty syringe weight, press return/enter. b. The VPP display should say ready to test c. Once the apparatus is ready press Start 15. The test is now under way 16. DO NOT TOUCH the apparatus/ test vessel etc while a test is running 17. Get the next sample test vessel ready by putting diluent in go to step 1 18. When the result appears write it down in the column for H2S

18

6 APPARATUS

The apparatus to be used for the robustness study will be:

No Instrument

H2S Serial Software No Issue 5.00 5.00 5.00 5.00 5.00 5.00 5.00 5.00

STANHOPE-SETA H2S Analyser SA 4000-0 1022108 2 3 4 5 6 7 8 STANHOPE-SETA H2S Analyser SA 4000-0 1022109 STANHOPE-SETA H2S Analyser SA 4000-0 1022111 STANHOPE-SETA H2S Analyser SA 4000-0 1022110 STANHOPE-SETA H2S Analyser SA 4000-0 1018696 STANHOPE-SETA H2S Analyser SA 4000-0 1022107 STANHOPE-SETA H2S Analyser SA 4000-0 1009089 STANHOPE-SETA H2S Analyser SA 4000-0 1009088

No Instrument

VPP Serial Software No Issue 1.11 1.11 1.11 1.11 1.11 1.11 1.11

STANHOPE-SETA VPP 1022205 2 3 4 5 6 7 STANHOPE-SETA VPP 1022206 STANHOPE-SETA VPP 1021374 STANHOPE-SETA VPP 1021373 STANHOPE-SETA VPP 1022208 STANHOPE-SETA VPP 1022310 STANHOPE-SETA VPP 1022209

19

STANHOPE-SETA VPP 1022207

1.11

20

7 EXAMPLE RESULTS PROFORMA

21

Appendix B
IP 570/12A - draft

Determination of hydrogen sulfide in fuel oils Rapid liquid phase extraction method
1 Scope

This document specifies a procedure for the determination of the hydrogen sulfide (H2S) content of fuel oils such as marine residual fuels and blend stocks, in the range ( to be advised) mg/kg with viscosities up to (to be advised) mm2s1 at 50 C, and marine distillate fuels in the range (to be advised) mg/kg as measured in the liquid phase. The method covers 2 procedures and relevant equipment.

NOTE 1 - By using a smaller volume of sample taken by a suitable positive displacement pipette, it is possible to extend the range of the instrument to over 200 mg/kg.

NOTE 2 - The precision of this method is critically dependent on the sampling, thermal history and handling of the test sample.

NOTE 3 - This standard is suitable for materials with a viscosity of over 3 000 mm2s1 at 50 C, (see 7.2.1), but the precision could be affected.

NOTE 4 - For the purposes of this standard, the terms "% (m/m)" and "% (V/V)" are used to represent respectively the mass fraction and the volume fraction.

NOTE 5 - Some samples can give elevated hydrogen sulfide readings if they contain thiols or alkyl sulfides above a liquid phase concentration of approximately 5 mg/kg. Elevated readings, as mentioned, are avoided by following Procedure A

WARNING 1 H2S is a very dangerous, toxic, explosive, colourless and transparent gas which can

22

be found in crude oil. It can be formed during the manufacture of the fuel at the refinery and can be released during handling, storage and distribution. At very low concentrations the gas has the characteristic smell of rotten eggs. However, at higher concentrations it causes a loss of smell, headaches and dizziness, and at very high concentrations can be fatal. It is strongly recommended that personnel involved in the testing for H2S are aware of the hazards of vapourphase H2S and have in place appropriate processes and procedures to manage the risk of exposure.

WARNING 2 The use of this standard may involve hazardous materials, operations and equipment. This standard does not purport to address all of the safety problems associated with its use. It is the responsibility of the user of this standard to establish appropriate safety and health practices and determine the applicability of regulatory limitations prior to use.

Normative references

The following referenced documents are indispensable for the application of this document. For dated references, only the edition cited applies. For undated references, the latest edition of the referenced document (including any amendments) applies.

IP 475, Petroleum liquids Manual sampling (EN ISO 3170)

ISO TR 13739, Petroleum Products Procedures for the transfer of bunker fuel(s) to ships

Principle

A weighed test portion is introduced into a heated test vessel containing diluent base oil. Air is bubbled through the oil to extract the H2S gas. The H2S is passed, with the air, via a filter cartridge which is held at -20oC (Procedure A only) to a detector to enable the H2S content of the air to be measured and the amount of H2S in the liquid phase to be calculated. In the case of Procedure B the filter cartridge is omitted.

Reagents and materials

Use only reagents of recognised analytical grade.

23

4.1 40 C.

Diluent oil1, proprietary water white API Group 2 base oil with a typical viscosity of 100 mm2/s at

4.2 Verification materials1, pressurized nitrogen, of at least 99,999 % (V/V) purity, containing certified levels of H2S.

4.3

Toluene, technical grade.

4.4

Petroleum ether (60/80), technical grade.

4.5

Acetone, technical grade.

4.6

Filter cartridge, see Figure A.4, individually packed in a sealed envelope

Apparatus

5.1

General

The apparatus, as detailed in Annex A, comprises an air pump, test vessel, heating jacket, filters, H2S specific detector, integral computer, automatic solenoid valves, gas flow detectors, a disposable syringe. and (Procedure A only) a vapour phase processor.

5.2 g.

Analytical balance, single pan or double pan balance capable of weighing to the nearest 0,001

5.3 1%.

Syringe, 20 ml polypropylene disposable type, for introduction of the diluent oil (4.1), accuracy

5.4 Pipette, 1 ml positive displacement pipette, for the introduction of the test portion. The accuracy as stated by the manufacturer should be typically 0,25 % at 1 000 ml, with a coefficient of variation of 0,04 %. It is recommended that the syringe be of the same type as specified by the equipment

A list of suppliers is available from the Energy Institute.

24

manufacturer for optimum system performance.

5.5

Disposable syringe, 5 ml, or 10 ml, for the introduction of the test portion. Typical accuracy

1%. This may be fitted with a needle or extra tube to allow sample to be taken from 3 cm below the surface, if appropriate.

5.6 Refrigerator (optional), for storing the test sample (see 6.4). The refrigerator shall be of a type suitable for storing volatile materials.

5.7 Oven/water bath (optional), for warming the sample to 40 C with an accuracy of 7.2.1). The oven shall be of a type suitable for use with volatile materials.

2 C (see

5.8 11.3).

Ultrasonic bath (optional), general purpose for cleaning the test vessel (note under clause

Sampling and sample handling

6.1

Unless otherwise specified, samples shall be taken as described in IP 475.

Care shall be taken to ensure that the integrity of the material is maintained and the possible loss of H2S is kept to a minimum. Where samples are drawn in a manner which does not minimise vapour loss (e.g. continuous drip sampling), dedicated samples for H2S determination shall be drawn.

6.2 Draw the sample directly into a suitable clean H2S inert container, of a minimum volume of 500 ml. The closure aperture shall allow the drawing of a test portion with the pipette (5.4) or syringe (5.5). To ensure sample integrity fill the sample container to approximately 95 % full and replace the cap immediately and securely.

NOTE 1 - It is recommended that containers such as dark brown borosilicate bottles or epoxy lined containers fitted with impervious gas-tight closures are used.

NOTE 2 - Lower volume containers may be used, however the precision could be affected.

NOTE 3 - In certain sampling situations it is impossible to draw samples directly into the sample container owing to the equipment that must be used, for example, devices used to take samples through the roof of

25

storage tanks, and closed system samplers such as those commonly used for ship compartments blanketed with inert gas. In these situations it is acceptable to transfer the sample from the sampling device into the sample container, however care should be taken to keep losses of H2S from the sample to a minimum during the transfer.

6.3 If Epoxy lined containers are used, these shall be visually inspected to ensure that the lining has not been damaged and that the containers are not dented.

6.4 Take the samples to the laboratory as soon as is practicable after sampling. Test immediately if possible. If samples are not tested immediately, store in a cool place such as a refrigerator (5.6) and analyse them within 3 days after sampling.

6.5 H2S measurement shall be the first test carried out on the sample as any additional handling can lead to loss of H2S and low results.

Sample preparation

7.1 To minimize the loss of H2S gas, do not homogenize, avoid shaking the sample, do not transfer the sample to another container and avoid unnecessary openings of the container before taking a test portion.

7.2 The sample needs to be flowing freely enough to allow it to be drawn into the pipette (5.4) or syringe (5.5).

7.2.1 A sample that is not free flowing at ambient temperature shall be gently warmed in a water bath or oven (5.7) set at a temperature not exceeding 40C.

NOTE It has been found that samples with a viscosity over 500 mm2/s at 50 C usually require warming.

Apparatus preparation

8.1

General

Follow the manufacturers instructions for the correct set up, verification, calibration and operation of the apparatus.

26

8.2

Location of apparatus

Use the apparatus under a suitable fume hood and vent the gas exit tube to a suitable extractor as the test will release small amounts of H2S gas during test portion introduction and during the measurement.

8.3

Filters

8.3.1

The inlet air filter shall be replaced every 3 months or if any discolouration is visible.

8.3.2 Follow the manufacturers instructions regarding the replacement intervals of the air pump and moisture filters.

8.3.3

Replace the moisture filter if any discolouration is visible.

Verification

9.1

Air flow rate

Verify that the air flow rate is 375 ml/min 55 ml/min at least once a year using a calibrated flow meter connected to the air output connection. If the flow rate is incorrect, recalibrate the flow (10.1).

9.2

Test vessel heater

Verify that the temperature of the heater is 60,0 C 1,0 C at least every 6 months, by inserting a calibrated platinum resistance thermometer into the heater. If the temperature is incorrect, recalibrate the heater temperature (10.2).

9.3

Validation of H2S detector

Follow the manufacturers instructions to verify the performance of the detector at least every month or when a new detector is fitted at zero using air, and at a nominal 25ppm vapor concentration

27

level using pressurized nitrogen (4.2) containing certified levels of H2S. If the measurement is not within 10 % of the certified value, replace the moisture filter and check the cleanliness of the liquid trap, test vessel and associated tubing. Purge the system with air for30 min and attempt the verification again. If the measurement is still outside the allowed tolerance, the detector needs recalibration or replacing. This verification is to be undertaken at ambient temperature.

9.4

Validate the temperature of the vapour phase processor (Procedure A only)

Verify that the temperature of the cooler is -20,0 C 2,0 C at least every 6 months, by inserting a calibrated platinum resistance thermometer into the cooler. If the temperature is incorrect, follow the manufacturers advice

10

Calibration

10.1

Air flow rate

Follow the manufacturers instructions to calibrate the air flow rate.

10.2

Test vessel heater

Follow the manufacturers instructions to calibrate the heater temperature.

10.3

H2S detector calibration

The detector is factory calibrated to allow for both the static and dynamic response to H2S. This calibration information is held digitally on the detector assembly and is read directly by the computer in the apparatus. The calibration enables results in mg/kg to be calculated from the test results.

Follow the manufacturers instructions for installing a new detector and verify the performance immediately afterwards (9.3).

11

Procedures A & B

11.1 General

28

It is recommended that Procedure A is used to determine the liquid phase Hydrogen Sulfide concentration. Test results obtained using Procedure B may be elevated by the presence of Thiols (Mercaptans) or alkyl Sulphides.

11.2 Procedure A

C 11.2.1 On switch on, the H2S apparatus shall control the test vessel heater to 60,0 by pumping air directly to the detector.

1,0 C and purge

11.2.2 Clean the test vessel and screw cap before each test. The test vessel shall be clean and dry before use. Fit the screw cap tightly.

11.2.3 Empty and clean the liquid trap if any liquid or discolouration is visible.

NOTE - Toluene (4.3) followed by petroleum ether (60/80) (4.4) and acetone (4.5), in conjunction with an ultrasonic bath (5.8) have been shown to be effective in cleaning the test vessel, cap, and liquid trap.

11.2.4 Insert a filter cartridge (4.6) into vapour phase processor, following manufacturers instructions

11.2.5 Initiate cooling on vapour phase processor, allow at least 10 minutes before proceeding with step 11.4.

Note an interlock between the analyser and vapour phase processor prevents the sample weight being entered before the filter cartridge has been cooled and is ready to test

11.2.6 Ensure the vapour phase processor has completed its cool down cycle before introducing the sample

11.3 Procedure B

11.3.1 Upon switch on, the H2S apparatus shall control the test vessel heater to 60,0 C purge by pumping air directly to the detector.

1,0 C and

11.3.2 Clean the test vessel and screw cap before each test. The test vessel shall be clean and dry

29

before use. Fit the screw cap tightly.

11.3.3 Empty and clean the liquid trap if any liquid or discolouration is visible.

NOTE - Toluene (4.3) followed by petroleum ether (60/80) (4.4) and acetone (4.5), in conjunction with an ultrasonic bath (5.8) have been shown to be effective in cleaning the test vessel, cap, and liquid trap.

11.4 All procedures (A & B)

11.4.1 Remove the screw cap from the test vessel and introduce 20,0 ml 0,5 ml of diluent oil (4.1) using the syringe (5.3) and replace the screw cap. Place the test vessel in the temperature controlled heating jacket and fit the input/output tubing.

11.4.2 Air is pumped through the diluent in the test vessel and to the detector for 5 min. This allows the diluent oil to warm up and the system to be purged.

11.4.3 Air is then pumped directly to the detector bypassing the test vessel to allow the test portion to be introduced without purge air affecting the concentration of the H2S. Ensure that the instrument is operating in bypass mode before introducing sample.Introducing the sample before this condition is met will likely result in premature loss of H2S and an erroneously low result. If this is suspected the test must be abandoned and repeated. The instrument will warn the user if they attempt to introduce the test portion before it is operating in bypass mode.

11.4.4 Depending on the expected H2S concentration, draw the appropriate volume of sample from at least 3 cm below the sample surface, avoiding the bottom of the container, into the disposable syringe (5.5) or 1 ml positive displacement pipette (5.4) and weigh to the nearest 0,001 g (5.2). Avoid applying a strong vacuum to the sample and take care to ensure that the integrity of the material is maintained so that the possible loss of H2S is kept to a minimum. Do not expel air from the syringe or pipette into the sample. Using the keypad, enter the total mass, which includes the pipette and sample, into the apparatus. The appropriate volume shall be determined by reference to Table 1. Table 1: Appropriate test volume for expected H2S concentration

Expected H2S concentration (mg/kg) 0 -10

Required test volume (ml) 5

Sample introduction

Disposable syringe

30

10 20 20+

2 1

Disposable syringe Pipette

Note A single weight can be entered rather than entering the weight before and after sample introduction if the balance is tared.

11.4.5 Introduce the test portion into the test vessel, ensuring that the syringe or pipette is held vertically to avoid sample adhering to the test vessel walls. Ensure that the syringe or pipette does not touch the surface of the diluent liquid. Any pickup of diluent oil onto or into the syringe or pipette will result in the mass of the oil being included in the mass of the empty syringe or pipette (11.4.7) leading to an erroneously high result. If this is suspected the test must be abandoned and repeated.

11.4.6 Immediately, as soon as the sample has been introduced to the test vessel, fit the screw cap tightly on to the top of the test vessel.

11.4.7 Weigh the empty syringe or pipette to the nearest 0,001 g (5.2) and using the keypad enter the mass into the apparatus.

11.4.8 Start the test. The following actions take place automatically:

11.4.9 The detector reading is normalized to zero.

11.4.10.1 In the case of Procedure A air is pumped directly through the filter cartridge for 3 minutes. After 3 minutes air is diverted through the test portion and diluent in the test vessel via the filter cartridge in the vapour phase processor and then carried to the detector

11.4.10.2 In the case of Procedure B air is pumped directly through the test portion and diluent in the test vessel and then carried to the detector

11.4.11 The millivolt readings (mV) from the detector are recorded at least every 4 seconds throughout the test until 15 minutes have elapsed. The result is automatically calculated and displayed in mg/kg.

31

11.4.12 Air is pumped to purge the detector.

11.4.13 Remove and clean the test vessel (11.3.2 11.3.3).

11.4.14 When the test vessel is removed for cleaning, check that the sample and diluent oil have properly mixed. If two distinct phases exist, check the gas connections and repeat the test.

12

Calculation

Calculate the H2S content of the sample, expressed in mg/kg, using the equation:

H2 S =

A xM m

where: A is the integrated area of cell output over the test time (mV.s).

is the calibration constant of the detector ( g)/(mV.s).

is the mass of the sample (g).

13

Expression of results

For all completed tests, record the following:

(i) H2S content to the nearest 0,01 mg/kg for < 10 mg/kg, and to the nearest 0,1 mg/kg for mg/kg. (ii) The procedure followed (A or B)

10

32

14

Precision2

14.1

General

The precision values for Residual fuels given in 14.2 and 14.3 were derived from a 2009 laboratory study which used 7 instruments and 7 operators at a single location, to test 15 samples in duplicate and in random order. The sample types covered the H2S levels of 0,01 to 15,3 mg/kg and viscosities up to 3 000 C. The precision values for Distillate fuels given in 14.2 and 14.3 were derived from a 2011 mm2s1 at 50 laboratory study which used 10 instruments and 10 operators at a single location, to test 8 samples in duplicate and in random order. The sample types covered the H2S levels of 0,40 to 9.70 mg/kg.

As the precision was determined from results obtained at a single location the reproducibility may not be comparable when results obtained at different times and locations are compared, due to sampling, shipping, storage and environmental factors. In practice two results obtained from different locations would be acceptable if their difference did not exceed the published reproducibility.

14.2

Repeatability3, r

The difference between two test results, obtained by the same operator with the same apparatus under constant operating conditions on identical test material would in the long run, in the normal and correct operation of the test method, exceed the following value in absolute value in only one case in twenty.

14.2.1 Procedure A

r = x.xxxxx

14.2.2 Procedure B

Residual fuels Distillate fuels

r = 0,2970 X 0,6 r = 0,2099 X 0,7

Laboratory studies were carried out at a single location

The round robin research report is available from the Energy Institute.

33

where X is the average of the results being compared, in mg/kg.

See Table 2 and Table 3 for a tabular illustration and Annex B for a graphical illustration of this relationship.

14.3

Reproducibility3, R

The difference between two single and independent test results, obtained by different operators working in different laboratories on identical test material, would in the long run, in the normal and correct operation of the test method, exceed the following value in only one case in twenty.

14.2.1 Procedure A

R = x.xxxxx

14.2.1 Procedure B

Residual fuels Distillate fuels

R = 0,5232 X 0,6 R = 0,2389 X 0.7

where X is the average of the results being compared, in mg/kg.

See Table 2 and Table 3 for a tabular illustration and Annex B for a graphical illustration of this relationship.

34

15

Test report

The test report shall contain at least the following information:

a)

a reference to this standard;

b)

the type and complete identification of the product tested;

c)

the result of the test (see clause 13);

d)

any deviations, by agreement or otherwise, from the procedures specified;

e)

the time and date of the test.

f)

the procedure followed (A or B)

35

Annex A (normative)

H2S Apparatus
A.1 Apparatus

A.1.1

General

The apparatus1, as shown diagrammatically in Figure A.1 (Procedure A) & A.2 (Procedure B), is self contained and operates automatically to measure the amount of H2S in liquid samples of fuel oils.

A.1.1.1 Air pump filter, 5 micron nylon encapsulated, to remove dust from the air.

A.1.1.2 Air pump, capable of supplying air at 375 ml/min

55 ml/min. The air is used:

a) b) c)

To purge the detector, glassware and tubing. To agitate the test portion and diluent mixture. As a carrier for the extracted H2S gas.

A.1.1.3 Inlet air filter, carbon type to remove moisture and contaminants from the air supply.

A.1.1.4 Solenoid valves, to divert the air supply from the test vessel as required during the normal test sequence and to divert the H2S gas away from the detector, if high levels of H2S are measured that could saturate the detector.

A.1.1.5 Flow sensors, electronic sensors incorporated to ensure that the air/H2S gas mixture reaches the detector, to enable the air flow to be controlled and that any flow problems are detected.

A.1.1.6 Liquid trap, 25 ml borosilicate glass tube with an internal input tube, to trap any liquid or heavy

36

vapours. The top half of the tube shall be loosely filled with glass fibre.

A.1.1.7 Moisture filter, 5 micron nylon encapsulated, to remove moisture from the air/H2S mixture.

A.1.1.8 Heater jacket, 50 Watt, aluminium block, controlled to 60,0 C cutout.

1,0 C, with an over temperature

A.1.1.9 Temperature probe, platinum resistance thermometer to measure and control the heating jacket temperature.

A.1.1.10 Test vessel, 50 ml borosilicate glass tube with an internal input tube for the air supply. This test vessel is heated by the heater jacket. See Figure A.2

A.1.1.11 H2S detector, electrochemical type specifically for H2S measurement, range (gas) 0-50 ppm (V/V), repeatability 1%, response time T90 < 30 seconds.

A.1.1.12 Vapour Phase Processor (Procedure A only), electrothermal device designed to cool the Filter cartridge (4.6) to -20oC and allow gases emitted from the test vessel to pass through the Filter cartridge and then to the H2S detector

37

Figure A.1: H2S apparatus with Vapour Phase Processor (Procedure A)

Figure A.2: H2S apparatus (Procedure B)

38

Figure A.3: Test vessel

39

40

Figure A.4 Sorbent Cartridge Key

  "!$#&%  '(!)01) 20120) 30 41#50 3 67 89A@BC51D3 150 %  13FEGH@IQP @"!$) 1R&% (S 3#50)41#5 3 TUIV W#&% 3

41

Appendix C

42

43

44

45

46

The following was the previous precision estimates with Sample 2 in place before it was realized that the whole sample should be removed. It is referred to in the letter accompanying the precision estimate above and hence is included. The precision estimate dated 16th February, 2012 is the one used.

47

48

49

50

51

52

Appendix D ! ( ( ) " #$% & '""

* ! + ,/ + ! ! 2 , * ( ( , + ! ! + * . '"", ! ( + ! ! + + (

" #$% ( !

* . * + + ! +

0, , +

+ 1 ++

! " # $ % & ' ( ( )* * +,. /! !% 0 10 /

53

+ ( 34 . ( ( + + 9 7 ! ! ; * ! , ! + + ! ! ! * (

* * , . !+ 1 4 0

) 1 + 4 , 5 2 , 7 ! + , ! ! + + + : ; + ! ! ! + ! , !0 2 * < + ! , * ! , 4+ * * 6 * + * . , * , + : %, + <! =4 + : ;, ( ! 0 !* + ! 2 8 , ! + . + * ! ( + + +

* ! + + ! > + + +

+ !

(A (

(B8 9C8" "8(B8D3 / 8B ( ! ! 4 1

54

! " ) + ( & , & ( * + ' + + / & ' . & * (( & ( , ,, & & ( & ( * , & ( $ (( * * ( . ' * ( $ + ( * & , & & & & & ' ' ) ( * ( * $ + * * ( ,, * (( ( & . $ ' ( ( $ ( * # (( ' $ ' , ' ) ) # $ % & ' ( ( * (( # $ * ( ( . * , ( . ( $

0 # 1 * & !45 #$ ( % ' 3 ; 2 , 6! 7* * (( $* & , * : & (( 3 & 5 >4$ 9 (( ' & (& :$ 3 * ( & 0:! 60; ( 3

8 & <,, * 8 40= &

55