Mercury Speciation and Distribution in an Egyptian Natural Gas

Oct 27, 2016 - The dominant Hg species found in the analyzed gas condensates were elemental Hg (Hg0) and inorganic Hg with the methylmercury fraction ...
0 downloads 0 Views 956KB Size
Subscriber access provided by University of Otago Library

Article

Mercury speciation and distribution in an Egyptian natural gas processing plant Mohamed Farouk Ezzeldin, Zuzana Gajdosechova, Mohamed B. Masod, Tamer Zaki, Jörg Feldmann, and Eva M. Krupp Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.6b02035 • Publication Date (Web): 27 Oct 2016 Downloaded from http://pubs.acs.org on November 4, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Energy & Fuels is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Mercury speciation and distribution in an Egyptian natural gas processing plant

Mohamed F. Ezzeldin,†‡ Zuzana Gajdosechova, † Mohamed B. Masod,‡ Tamer Zaki,‡ Jörg Feldmann, † Eva M. Krupp†* †

Trace Element Speciation Laboratory, Department of Chemistry, Meston Walk, University of

Aberdeen, Aberdeen, AB24 3UE, UK ‡

Egyptian Petroleum Research Institute (EPRI), Nasr City, 11727, Cairo, Egypt

ABSTRACT: Unprocessed petroleum hydrocarbons often contain high concentrations of mercury (Hg), which can severely damage the metal components of a processing plant and pose a health risk to the workers and the natural environment. While Hg removal units can significantly reduce the Hg concentration in the export products, they are often installed in the final stage of the processing plant, thus failing to protect the production facilities. In this study, Hg distribution within a natural gas processing plant was studied in order to identify the most effective place for an Hg removal unit. Additionally, impact of sampling container materials and their acidification was evaluated and Hg species in the condensate were quantified. Total Hg concentration was significantly higher in all samples stored in the glass in comparison with plastic containers. However, the acidification effect of the containers was more pronounced for Hg in non-polar solutions. Interestingly, the assessment of Hg distribution within the gas plant showed that the export gas is being enriched in Hg, which concentration has risen from 1.25 to 4.11 µg/Sm3 during the processing steps. The 2nd stage separator was identified as the source of excess Hg, which partitioned from the liquid phase of condensate to

1 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 22

the gas phase as a result of reduced operational pressure and temperature. The dominant Hg species found in the analyzed gas condensates were elemental Hg (Hg0) and inorganic Hg, with the methylmercury fraction comprising up to 18%. However, it was also found that the % fraction of individual Hg species varied along the plant units most probably as a result of Hg0 migration to the export gas. Therefore, in order to protect all treatment facilities from Hg contamination, the Hg removal unit should be installed after the 2nd stage compressor.

INTRODUCTION Mercury naturally occurs in oil and gas reservoirs, however its concentration in petroleum hydrocarbons is highly dependent on the geological location and varies between 0.01 and 10,000 ppm.1 Individual Hg species show different chemical behavior such as adsorption, amalgamation, solubility, vapor pressure, and phase partitioning (gas/liquid hydrocarbons/water) during treatment in a natural gas plant. In the case of mixed phases, organic Hg species preferentially partition to liquid fractions (condensate) and ionic compounds escape to produced water, while Hg0 equilibrates between liquid and gas fractions.2, 3 However, quantitative data for Hg and Hg species within a gas production plant are extremely scarce. Mercury impacts on gas processing plants as well as refinery industries in several ways.4 In natural gas treatment plants, separation of ethane and heavy hydrocarbons from sales gas, mainly methane, depends on a cryogenic distillation process using cold boxes. While in the refinery, the components of natural gas liquids (NGL), which are collected by cryogenic distillation, are separated through several thermal fractionation steps based on boiling points.5 In both treatments, a heat exchanging process is applied, typically using aluminum boxes to facilitate heat exchange. Unfortunately, Hg can easily degrade the aluminum (Al) heat exchanger boxes through three basic mechanisms: amalgamation, amalgam corrosion and liquid metal embrittlement (LME).6 During amalgamation processes, Hg0 forms liquid solutions with Al, however as the concentration of Al in the Al-Hg amalgam is relatively low, defects caused by this mechanism are generally shallow in depth.6 On the other hand, amalgam corrosion propagates even with miniscule amounts of Hg in the presence of moisture. During this process, Al within Al-Hg amalgam is rapidly oxidized by water thus removed from the amalgam. Al depleted amalgam therefore dissolves further Al and this

2 ACS Paragon Plus Environment

Page 3 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

process continues until all Al is completely oxidized.7 However, as the presence of water in the Al heat exchangers is very unusual, this type of Hg induced defects are very rare. Liquid metal embrittlement is thought to be the most common Hg induced degradation, which can severely damage the Al heat exchangers. During this mechanism, Hg atoms adsorb at stressed surfaces and crack tips usually along grain boundaries, which weaken substrate interatomic bonds.8 Once cracks have been initiated, sub-critical cracking can occur very rapidly (>100 mm/s) even at low stress.9 However, the most important condition for Hg induced degradation of the Al box is defect formation in the oxide layer (Al2O3), which is protecting Al alloys from the direct contact with Hg. The thin oxide layer can be broken mechanically via abrasion of the surface by small particles present in the gas stream.7 However, as Al boxes are under constant thermal stress it is possible that the thermal expansion differences between Al alloy and the oxide layer may cause the layer to crack when the heat exchanger is warmed.6 Thus Hg can cause severe corrosion and cracking of the Al heat exchangers10 and there have been several cases of heat exchanger failure recorded in natural gas processing plants and ethylene production plants worldwide with several devastating consequences, as e.g. the Skikda gas plant explosion.11 Furthermore, Hg present in petroleum hydrocarbons poisons the catalysts and increases the exposure risk of the field workers and the immediate environment. In this study, we present a comprehensive quantification of Hg in an Egyptian natural gas processing plant with the aim to identify the most suitable position for introduction of an Hg removal unit, which would help to reduce Hg contamination in the petroleum hydrocarbons and simultaneously prevent possible damage to the processing plant. Therefore, samples of natural gas, gas condensate, produced water, Benfield solution and TEG solution were collected before and after each processing unit and analyzed for total Hg and Hg species (Hg0, iHg, MeHg, Me2Hg). Additionally, specific attention was paid to the suitability of sampling containers and the quantification accuracy of identified Hg species.

EXPERIMENTAL Reagents and Standards Stock standard solutions for Hg and thallium (Tl) at a concentration of 1,000 ± 3 mg/L in 12% HNO3 acid and sulphuric acid (H2SO4) for trace analysis were purchased from Fluka (UK). High

3 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 22

purity nitric acid (69%) was obtained from BDH (UK). 36% hydrochloric acid laboratory reagent was purchased from Fisher Scientific (UK). Tin chloride dihydrate (SnCl2.2H2O) 98%, methylmercury(II) chloride (MeHgCl) and isooctane (2, 2, 4-trimethylpentane) 99% were obtained from Sigma-Aldrich (UK). Grignard reagent (BuMgCl) in 2 M THF was purchased from Sigma Aldrich (Germany). Deionized water was obtained from a laboratory bi-distillation purification system Aquatron A4000D (UK).

Description of the natural gas plant The streams produced from the gas wells are gathered by the field-gathering network and directed towards the plant.12 The treatment strategy relies on dividing the incoming stream into two parallel trains (tr-1 and tr-2) and each train includes an inlet separator (phase separation separator), acid gases removal unit, dehydration unit and LTS (low temperature separator), or commonly named Al heat exchanger. The incoming stream enters a horizontal inlet separator to separate gas, gas condensate and water from the mixed stream. The sour gas (H2S rich gas) goes through a sweetening process using hot potassium bicarbonate (Benfield solution) at about 110 ºC, to remove H2S. The acid-rich Benfield solution is recycled and the sweetened gas (of low H2S concentration) is directed towards the dehydration unit, which contains both a TEG (triethylene glycol) contactor and a re-generation package to remove trace amounts of water dissolved in the gas. The dry gas flows from the TEG contactor into the LTS (heat exchanger) to separate traces of heavy hydrocarbons remaining in the gas, and the rich TEG is regenerated. Afterwards, the processed natural gas is fed into the Egyptian national pipeline network under pressure of 100 bar, whilst the separated heavy compounds are directed towards the NGL (natural gas liquid) unit. According to the design of this plant, the stabilization of gas condensate is performed in two successive steps, in the 2nd stage separator and in the stabilizer tank. The separated gas stream is routed via a common header to the 2nd stage recycle compressor before being sent to the sweetening process while the condensate is redirected towards the storage tank. Export condensate is moved from the storage tanks via shipping pumps and the produced water collected from both the inlet

4 ACS Paragon Plus Environment

Page 5 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

separator and the 2nd stage separators is disposed of. A schematic illustration of the plant with a detailed description of the production process is available in the Supporting Information, Figure S1.

Samples and collection points All samples were collected by Egyptian petroleum research institute (EPRI) with permission from the Ministry of petroleum and mineral resources (Egypt) and delivered to the TESLA laboratory (University of Aberdeen, Aberdeen, UK) for total Hg and Hg speciation analysis. Liquid samples of condensate, produced water, Benfield and TEG solutions were collected from all available sampling points during normal processing conditions. The samples were collected directly from the taps, to avoid Hg loss or adsorption, which might be possible when using a connecting tubing, and valves were flashed for approximately 10 min for equilibration and to remove any precipitated solids that might have been present inside the valve. The individual collection sample points, pressures and temperatures are given in Table 1.

Table 1.

Sample collection and sampling containers Pre-cleaned 40 mL amber glass bottles (according to US EPA Protocol B13) were supplied from Supelco (UK) with a certificate of analysis. The bottles were assembled with type 1 borosilicate glass, PTFE/silicone septa, and a polypropylene cap. High quality, translucent and sturdy 125 mL polyethylene bottles (PE) were purchased from Supelco (UK). The autoclavable sampling bottles suitable for hot liquids sampling were equipped with a sealing cone in the screw-cap and narrowmouth to minimize possible evaporation of the sample. Natural gas, condensate, process water, TEG and Benfield solutions were sampled before and after the treatment units allowing for identification of Hg partitioning and enrichment within the processing areas of the natural gas plant. The flow rates of condensate feed form the train 1 and 2 were 15 and 13.7 m3/h, respectively, whereas the natural gas flow rates were 182 and 183 million standard cubic feet per day (MMSCFD. The export natural gas flow rate was 365 MMSCFD. Additionally, individual sampling containers, glass and PE

5 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 22

bottles were subdivided for sampling in acidified and non-acidified containers. Acidification of the sampling containers was done by addition of 0.5 and 1.25 mL of HNO3 into 40 and 125 mL sampling containers respectively after the sample collection. The acidified mixture was manually shaken for 2 min.

Total Hg analysis in natural gas A WA-4 mercury analyzer (Nippon instruments Co., Japan) was used to determine Hg0 concentrations in natural gas samples during 3 successive days within the same time period as the collection of liquid samples was undertaken. The analyzer was equipped with a low pressure Hg discharge lamp as light source at a wavelength at 253.7 nm and two programmable thermal desorption units (heating furnaces) capable of heating quickly to 700 °C and 800 °C. The furnaces were attached to two collector tubes (gold-coated sand trap) and the samples from the highpressure pipeline (up to 100 bar) were collected using a specially designed high-pressure sampling box (M&C Co., Germany), to reduce the pressure without changing the Hg concentration. The sampling and analysis of gas were done according to the standard method.14 Calibration of the analytical system was done by injection of known amounts of standard gas saturated with Hg vapor using a Nippon mercury standard gas box model MB-1 (Hg STD Gas Box). The box contains a drop of pure metallic Hg, which vaporizes to produce saturated Hg vapor, which is dependent on the ambient temperature, thus the concentration of Hg0 in the gas can be determined. Four ranges of Hg detection levels, namely 0-2 ng, 0-20 ng, 0-200 ng and 0-1000 ng were applied according to Hg concentration in gas samples.

Total Hg analysis in gas condensate, process water, TEG and Benfield solutions For total Hg determination in the different liquid samples, prior to aliquotation, the samples were manually shaken for 10 min to homogenize. Accurately weighed amounts of approximately 1 mL of sample were mixed with 5 mL of HNO3 (70%) in 22 mL glass vials with Teflon-lined caps, and left for pre-digestion overnight (Gas condensate and TEG). Benfield solutions and produced water were predigested for 120 min. The pre-digested samples were then digested in an autoclave at 90 °C for 150

6 ACS Paragon Plus Environment

Page 7 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

min. The resulting digest was allowed to cool to room temperature and stored at 4 °C until further use. Total Hg measurements were done using CV-AFS (Millennium Merlin, 10.023, PS Analytical Ltd, Kent, UK). The instrument was calibrated with Hg standards at 0, 0.1, 1, 10 and 100 ng/g. Briefly, the digested samples were mixed with 5% HNO3 and 2% SnCl2 in HCl (1.1 M) in a mixing valve, where Hg2+ is reduced to Hg0 and swept out of solution by a stream of argon gas (0.3 L/min) and dried via a hygroscopic ion-exchange membrane. A high intensity Hg lamp with a wavelength of 253.7 nm was used as an excitation source for Hg measurements, and the fluorescence of Hg in the sample is detected.15 Deionized water and 0.1% mercaptoethanol solution were used for decontamination of the system.

Hg species analysis An HP-6890 Gas Chromatograph (Agilent Technologies, Santa Clara, CA, USA) was hyphenated via an in-house built heated transfer line with an Agilent 7500 ICPMS (GC-ICP-MS; Agilent Technologies, Santa Clara, CA, USA) was used for speciation analysis. Optimised analysis parameters can be found in Table 2. Prior speciation analysis, samples were manually shaken for 10 min. Each gas condensate sample was directly injected into the GC-ICP-MS without any dilution, to identify any volatile, dialkyl Hg species. For analysis of non-volatile mercury compounds, butylation with Grignard reagent was used. Accurately weighed amounts of approximately 1 mL of a sample were transferred into septum capped 22 mL pre-cleaned glass vials, 200 μL of 2 M Grignard reagent was added and the samples left to react for 5 min. Then, 10 mL of 0.5 M H2SO4 was added to stop the reaction. Finally, the entire mixture was centrifuged to facilitate separation of the organic and aqueous layer for 10 min at 3747 x g. The organic layer was transferred into amber GC vials and stored at -20 °C until analysis. Using this procedure, a calibration (0, 1, 10 and 100 ng/g) was established for MeHgBu and Bu2Hg, with retention times for MeHgBu as 210 s and 280 s for Bu2Hg, with R2 > 0.9999.

Table 2.

7 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 22

Analytical performance The analyses of all samples were done in triplicate, and each digestions cycle contained at least one blank, which was treated in the same way as the samples. The detection limit for total Hg analysis was calculated as 3 times the standard deviation in Hg concentration of 10 blanks, corresponding to 0.08 ng/g in the condensate sample, 0.05 ng/g in TEG solution, 0.01 ng/g in produced water and Benfield solution. The detection limits for Hg species were determined by three times the standard deviation of the background noise close to the chromatographic peak and resulted in LODs for inorganic Hg (iHg) of 0.11 and methylmercury (MeHg) of 0.13 ng/g.

RESULTS AND DISCUSSION Effect of sample container type and acidification on total Hg concentration

Figure 1.

Although more Hg was found in the glass containers compared to plastic containers, the material effect was less significant than the acidification, especially for the gas condensate in comparison with TEG solution (Figure 1 and Figure S2, S3). However, while total Hg concentration in the produced water and Benfield solution was found to be higher in the samples kept in the glass bottles, the effect of acidification was not significant (p = 0.51 and 0.69) (Figure 1 and Figure S4, S5). The observed differences in total Hg concentration between samples kept in the acidified and nonacidified glass bottles may be attributed to individual Hg species present in these samples and the polarity of the sample matrix. Glass tends to be negatively charged and would attract positively charged or MeHg species and therefore, adsorption on the walls of the glass container will lead to decrease of these Hg species in the matrix. However, acidification of the glass surface results in formation of neutral or positively charged layer, which leads to repulsion of Hg species from the container surface and thus helps to stabilize them within the matrix. Interestingly, the acidification

8 ACS Paragon Plus Environment

Page 9 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

effect of the surface has been seen diminished when aqueous solutions was added, hence the acidification effect was expected to be larger for non-polar solutions. Bloom16 performed two independent experiments to examine the stability of Hg0, inorganic, dimethyl- and methylmercury chloride species in both paraffin oil and composite crude oil sample for 25 and 60 days respectively. The initial Hg concentrations were 90 - 650 ng/g for spiked paraffin oil, while it was 200 ng/g for spiked oil composite. It was found that iHg exhibited significant loss from the spiked paraffin oil, when stored in a glass container due to the adsorption on the container walls. But interestingly, the loss of iHg was not noticed in the composite oil matrix, which most probably enhanced the stability of this Hg species. On the contrary, Hg0 levels were dramatically decreased in the composite oil matrix, whereas its stability level appeared to be enhanced by the paraffin oil. While the composition of both matrices is almost the same, there is a clear difference in the stability towards Hg species. Although, there is an understanding for packing and storing the sample without a headspace in the above experiments (to avoid evaporation and/or species transformation), health and safety regulations forbid to collect, store or transport petroleum samples without a headspace to permit the expansion of the samples and consequently avoid an explosion risk. Certainly, that will directly affect the Hg speciation process, making it a challenge for the analyst charged with analyses of oil matrix for volatile species such as Hg0. This has been confirmed by Snell and coworkers,17 who identified significant loss of Hg0 and HgCl2 in organic matrix over time.

Hg distribution through different processing stages of the Egyptian natural gas plant The processing operation of the Egyptian natural gas plant depends chiefly on three cycles; gas, gas condensate and produced water. Since the analysis of all matrices collected from the different processing stages in the natural gas plant showed significant Hg contamination (Figure 2, Table 3), it was important to study Hg distribution throughout the different parts of the processing facility, in order to propose the best location to install the Hg removal unit inside the plant. Installation of the Hg removal unit after LTS is able to minimize Hg contamination in the export products, however most parts of the processing plant would be under constant exposure to Hg. Thus, it is essential to

9 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 22

identify processing stage, at which removal of Hg would have a significant impact on its reduction through the production facilities.

Table 3. As shown in Figure 2 and Table 3, the natural gas contained 1.25 µg of Hg /Sm3 right after the inlet separator, while the Hg concentration in the export gas had risen to 4.11 µg/Sm3. This means that the natural gas fraction was enriched in Hg through the processing steps within the plant. It is apparent that the Hg enrichment process takes place in the sweetening step, as its concentration in the sweet natural gas was 11.6 µg/Sm3 while the vent gas from the sweetening process contained only 0.77 µg of Hg/Sm3. However, to understand where the access Hg originated, other production processes needed to be analyzed. Within the condensate cycle, a huge amount of Hg was lost in the 2nd stage separator. The separated condensate from the inlet separator contained 1117 ng/g of Hg but the outlet concentration was found to be only 43.6 ng/g. A further decrease in Hg concentration was noticed in the outlet condensate of the stabilizer unit (31.2 ng/g), and consequently a concentration of 26.7 ng/g was reported in the exported condensate. The outlet water from the inlet and 2nd stage separators were the only sources of produced water in the plant. Hg concentrations in produced water coming from the inlet and the 2nd stage separator were 31.2 ng/g and 37.7 ng/g, respectively. In contrast to the condensate, no significant difference was noticed in Hg concentration between the two separators. Using the recorded flow rates of the individual petroleum hydrocarbon streams, we were able to calculate the exact amount of Hg circulated through the plant at the time of sampling. The combined inlet stream contained 163.98 g of Hg/day, which was subdivided between the gas and gas condensate feed with each containing 89.19 g/day and 74.79 g/day, respectively. The amount of Hg in the export natural gas dropped to 51.88 g/day and in the export gas condensate to 1.99 g/day, indicating that over 110 g of Hg day-1 was released to the environment through gas vents and produced water.

Figure 2.

10 ACS Paragon Plus Environment

Page 11 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Interchange process Three sources were considered as the feed to the 2nd stage compressor; the outlet gases of the 2nd stage separator (37.2 µg/Sm3), the stabilizer unit (7.06 µg/Sm3) and NGL (not measured in this part of the experiments due to a technical fault), while the outlet from the 2nd stage compressor has a Hg concentration of 41.6 µg/Sm3. Hence, the Hg concentration in the outlet gas from the stabilizer unit provided the highest feed to the 2nd stage compressor, while the concentration from its outlet was the highest Hg concentration of all analyzed gas points (41.6 µg/Sm3). Consequently, the high Hg concentration in the feed gas from the 2nd stage compressor to the Benfield solution explains the unexpectedly high Hg concentration within the Benfield units, while the large loss of Hg concentration in the condensate of the 2nd stage separator contributes to the high Hg levels in its outlet gas. Such a phase transfer of Hg may be assisted by the decrease in both pressure and temperature inside the 2nd separator, which might also inflict changes on Hg species. Therefore, it could be proposed that natural gas reflects the Hg concentration of condensate. Similar results were reported by Petronas in a study on an offshore Malaysian natural gas processing plant in 2007.18 Additionally, it is likely that Hg0 was the dominant Hg species in the incoming stream from the wellhead due to its high volatility and ability to transfer from one phase to another. Furthermore, the occurrence of high Hg concentrations in the vent gas of the TEG re-generation unit (377 µg/Sm3) could be explained by the considerable solubility of Hg in the glycol solution as previously reported.19 The treatment of Hg in the condensate stream should be given top priority as suggested by the Petronas plant.18 In contrary, here we recommend the installation of a Hg removal unit after the 2nd stage compressor in this gas plant to protect all treatment facilities from Hg contamination. Additionally, attention should be paid to the Hg-contaminated vent gases, especially from the TEG re-generation unit, as they must be considered continuous sources of Hg exposure for plant workers. Furthermore, treatment of the discharged wastewater should also be targeted to protect the natural environment.

Hg speciation in gas condensate samples collected in the natural gas plant

11 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 22

Analysis of Hg species was performed in two steps; firstly, Hg0 was determined by direct injection (no other volatile Hg species were detected) and subsequently iHg and mono-alkyl Hg (RHg) species were determined following Grignard derivatization. Such a two-step analysis would also allow for detection of Me2Hg or other di-alkyl Hg species present in the samples. The main Hg species in the studied gas condensate samples were Hg0, iHg and MeHg (Table 4). All gas condensate samples were contaminated with these three Hg species, but their individual concentration differed considerably.

Table 4.

Hg0 and iHg were the dominant species in the gas condensate samples, which is in agreement with previously reported analysis.20-22 MeHg was also detected, but in low abundance in comparison to other species, up to 18%. It is worth mentioning that several previous studies reported the occurrence of MeHg in condensate and crude oil samples, however their concentration was always negligibly low.3, 16 The collection origin of the samples analyzed can be divided into two parts: the inlet part, which includes the collected samples from inlet separator trains 1 and 2 and samples from the incoming stream from the well head; the processing part, which includes samples from the 2nd stage separator, stabilized condensate and the export condensate; and finally samples from the different treatment stages within the plant. In the first part, Hg0 was the dominant species where its abundance was in the range of 85% to 89% in the condensate samples (Table 4). Additionally, the abundance of iHg was in the range of 10% to 13%, whilst the abundance of MeHg was calculated to be 2%. In contrast to the inlet part, iHg was reported to be the dominant species in all samples coming from the processing part with abundances ranging from 50% to 66%. Additionally, the abundance of MeHg was found to be significantly higher in the stabilized and export condensate, in the range of 9% to 18% and with Hg0 abundance falling to 25 - 32%. The significant changes in the abundance of these Hg species could be explained by the decrease of Hg0 concentrations throughout the various treatment stages. The loss of Hg0 from processed condensate during the 2nd stage and stabilized separators may refer to partitioning of Hg0 between

12 ACS Paragon Plus Environment

Page 13 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

natural gas (gas phase) and gas condensate (liquid phase), which is more volatile than the iHg and MeHg species, to the gas phases (natural gas). The chloride salts of Hg (HgCl2 and MeHgCl) are known to be more volatile than Hg0 as indicated by their vapor pressures (Hg0 = 0.26 Pa at 25 °C,23 MeHgCl = 1.13 Pa at 25 °C24 and HgCl2 = 100.5 kPa at 28.85 °C25). However, here iHg and MeHg species seem less volatile than Hg0, which may reflect that they bind to other ligands, e.g. containing sulfides (potentially H2S or COS), which may decrease the volatility of Hg species present in the studied samples. Consequently, these data support the assumption already observed in decreased total Hg concentration within condensate samples from the 2nd stage separator and stabilizer, which was accompanied with an increase in the Hg level in natural gas (Figure 2).

Figure 3.

It should be noted that the sum of Hg species accounts for only a fraction of the total Hg concentration analyzed by CV-AFS (Table 4). This observation was particularly evident in the samples from the inlet part (47 - 58%), which contained high concentrations of Hg0. Comparison between the concentration of Hg0 in the condensate samples determined before and after Grignard derivatization showed loss of Hg0 when derivatization was performed (Figure 3). The loss in the Hg0 concentration for all condensate samples was 18% to 75% and may have occurred during the sample preparation step (derivatization) as was previously reported.10 Additionally, the loss of Hg0 can also occur during the opening of the sampling bottles to handle aliquots, which should be kept to a minimum.

CONCLUSION The results presented in this work demonstrate that both natural gas and condensate are highly contaminated with Hg in the studied Egyptian natural gas plant. Our work suggests that there is significant loss of Hg from the samples stored in plastic containers either due to adsorption on or diffusion through the walls of the container. Acidification of the sampling containers improved storage capabilities of Hg present in non-polar matrices. In addition, Hg concentration in the export gas was approximately 4 times higher than the concentration in the inlet gas after the first separator. Significant Hg enrichment took place in the 2nd stage separator due to Hg partitioning to the gas phase from liquid phase, induced by reduction of operational temperature and pressure. Therefore,

13 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 22

in order to protect processing facilities from Hg induced defects, the Hg removal unit should be installed after the 2nd stage compressor. While Benfield solution showed low Hg content, TEG solution contained significant Hg concentrations due to the high solubility of Hg in glycols. The vent gases, especially those coming from the TEG cycle, contained an extremely high concentration of Hg and, therefore, their treatment is highly recommended. Although water samples contained the lowest Hg concentrations, its discharge should also be treated.

Table 1. Description of collected samples and sampling conditions from the natural gas plant.

Temp. Sample

Pressure

Description ( C)

(bar)

o

Condensate Inlet Sep, tr-1

Condensate from the outlet of inlet separator train-1

69.3

70.2

Inlet Sep, tr-2

Condensate from the outlet of inlet separator train-2

70.7

69.7

2nd stage Sep

Condensate from the outlet of 2nd stage separator

63.3

23.5

Stabilized

Condensate from the outlet of stabilizer tank

150

9.36

Export

Condensate from the outlet of export tank

21.3

5.12

Inlet Sep, tr-1

Water from the outlet of inlet separator train 1

69.3

70.2

Inlet Sep, tr-2

Water from the outlet of inlet separator train 2

70.7

69.7

2nd stage Sep

Water from the outlet of 2nd stage separator

63.3

23.5

Lean, tr-1

Benfield solution from the inlet of train-1 absorber

112

0.61

Rich, tr-1

Benfield solution from the outlet of train-1 absorber

102

69.2

Lean, tr-2

Benfield solution from the inlet of train-2 absorber

110

0.8

Rich, tr-2

Benfield solution from the outlet of train-2 absorber

102

68.4

Water

Benfield sol.

14 ACS Paragon Plus Environment

Page 15 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

TEG sol. Lean, tr-1

TEG solution from the inlet of train-1 absorber

65.2

70.6

Rich, tr-1

TEG solution from the outlet of train-1 absorber

51.6

68.2

Lean, tr-2

TEG solution from the inlet of train-2 absorber

65.5

70.6

Rich, tr-2

TEG solution from the outlet of train-2 absorber

53.2

68.3

Table 2. Optimized GC-ICPMS parameters

GC parameter

Optimized value

Injection type, volume,

Splitless, 1 µL,

Injector temperature

280 °C

Column

100% PDMS DB-1 (30 m x 0.59 mm x 1.0 µm )

Carrier gas

Helium, 10 mL/min

Makeup gas

Argon, ~300 mL/min

GC program

50 °C (1 min) / ramp 50 °C/min / 280 °C (1 min)

Transfer line

Inner: MXT guard column, i,d 0.28 mm, o.d. 0.53 mm Outer: Silcosteel, i,d 1.0 mm, o.d. 1/16 inch 0

Transfer line temperature

220 C

ICP-MS parameter

Optimized value

Plasma gas flow

Argon, 16 L/min

Make-up gas flow

Argon, ~300 mL/min

Carrier gas flow

Argon, ~950 mL/min

Spray chamber

Cyclonic

Nebuliser

Micro-Concentric Teflon PFA; ~300 µL/min

Internal standard

Continous aspiration Tl, 25 µg/L in 1 % HNO3

Dwell time

40 ms for 10 ms for

200

203

Hg,

Tl,

202

205

Hg;

Tl

15 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 22

Table 3. Distribution of total Hg concentration along the operational units of the natural gas plant.

Matrix

Location

Total Hg*

Natural gas

Inlet separator

1.25 ± 0.39

Benfield absorber

11.6 ± 3.04

TEG absorber

9.61 ± 1.88

LTS unit – export

4.11 ± 0.29

2nd separator

37.2 ± 9.65

Stabilizer

7.06 ± 2.01

2nd compressor

41.6 ± 7.66

Inlet separator

1117 ± 42.9

Condensate

nd

2

Benfield solution

TEG solution

Produced water

53.6 ± 1.56

Stabilizer

31.2 ± 1.28

Export

26.7 ± 1.22

Benfield absorber – in