Produced Water (PW) is the water associated with oil during the extraction process [1] . In oil and gas formations, the subsurface rocks are mostly filled with fluids such as water, oil and/or gas, or some combination of these three main components. Thus, oil and gas reservoirs often contain both formation water as well as hydrocarbons [2] . PW includes a number of water-soluble and insoluble organic and inorganic compounds [3] , and is considered one of the largest streams of generated wastewater. The global annual production of produced water is estimated to have reached 250 million barrels per day in 2020 [4] . Amongst the compounds found in PW are inorganic salts and a wide range of oil related hydrocarbon compounds. The type and concentration of metal and metalloid salts in PW vary widely and depend on the age and geology of the oil and gas field [5] . It is important to treat the produced water associated with the crude oil operations, where this water upon reinjection into the production zone may affect the quality of the oil, its properties, and therefore its price [6] . If not re-injected but disposed of, environmental considerations are of primary concern. In this regard, purifying the produced water to such a degree that it can be re-injected into the production zone or that it can be utilized for another purpose rather than stored, contributes to sustainable resource management by reducing the generation and accumulation of waste [7] . In our study of possibilities of purifying produced water from a Kuwaiti oil production site, we initially studied the impact of the produced water on the operation, focusing on potential damages to the subsurface equipment, specifically the production tubing, which carries the production fluid from the subsurface to above-ground. Corrosion of the tubing that carries the production fluid is a major worry [8] - [14] when, evaluating the structural integrity of the oil production facility. It has been noted that as many as 33% [15] to 51% [16] of oil/gas pipeline failures are caused by corrosion of the pipes [13] [14] . It has been reported that in the case of produced water from sour oil/gas fields tubing, failures are frequent and the workover rate significant [8] .

The focus of this study is an oilfield located in South Kuwait. It is a mature field with a water cut of 1:4. Production from 40 - 60 different wells is combined to supply a central processing substation (Figure 1). The total field is comprised of over 1000 wells.

The tubing in use in the subsurface is made of steel (for details, please consult Table 1). A typical length of tubing used from the active production zone to the central process unit is in the range of 300 m - 675 m, with a total tubing for the complete operation connected to one substation of about 40.000 m (Figure 2).

The scale formation in pipe tubing in oil and gas fields is influenced by a number of factors such as the nature of the produced water, the temperature [17] , pressure [18] , pH [19] , and flow velocity [20] [21] . Due to its high ion content, especially the high chloride [22] , but also calcium content [23] , produced water is corrosive to metals. In addition, produced water from sour oilfields often has a relatively low pH [24] which again supports metal corrosion [25] . Because of this reason, 50% of tubing has to be exchanged every 6 months at a cost of $12.9 per tubing meter (Figure 3). The corrosion adds to non-soluble particles in the produced water and leads to additional scaling within the tubing. The scale and residual material block the production tubing’s inner space, which prevents the oil from flowing from the production zone to the surface (Figure 4). This causes a drop in the well’s productivity and requires a workover job to be performed on the well to restore its production capacity.

The current contribution looks at the impact of the produced water on the production tubing, investigates the composition of the deposits on the inner surface of the tubing, and the composition of the scaling. The work focuses exclusively on the subsurface tubing and not on the tubing above ground. The tubing above ground has different characteristics than that in the subsurface, both in steel composition and dimensions, including the thickness (9.5 mm above-ground vs. 5.7 mm in the subsurface), that makes the tubing above ground less prone to fail. Some of the questions that were looked at in this study

were what contribution the metal debris from the pipes would make to the overall scaling, what iron compounds would be included in the scaling, whether sulfidic iron compounds would be included in the scaling and whether one could observe evidence of chloride induced corrosion.

Produced water samples, tubing, including connection pieces and scaling samples [in this contribution, solids that have deposited on the surface on the tubing as well as debris stemming from the tubing that has lodged itself in the tubing is seen as scaling] were sourced from a production site in a sour oilfield in South Kuwait. The produced water analysis was carried out at the CORE labs, Abu Dhabi, UAE. The surface textures of the scaling samples and the tubing were investigated with a stereo microscope (Model SZ2-ILST) and a scanning electron microscope (SEM, Thermo Quattro S). As the raw scalings were coated with hydrocarbons and other carbonaceous material, raw scalings were submitted to “ashing” at 600˚C in a Carbolite oven, in order to evaluate the inorganic composition, including the metal content and metal salts. Also, raw scalings were extracted with toluene at rt. Regarding visual textural analysis of scalings and tubing, photos or micrographs were taken with a stereoscope (Model SZ2-ILST) and a scanning electron microscope. Regarding spectroscopic analysis, Fourier-Transform infrared (FT-IR) spectroscopy of the samples was carried out with a Perkin Elmer Spectrum 2 FT-IR spectrometer. WD-XRF spectra on the samples were performed with Rigaku Primus IV. XPS spectra on the samples were carried out with a Nexsa G2 (Thermo Scientific, UK). In addition, SEM was performed in combination with energy dispersive X-ray spectroscopy (EDS).

Collection of produced water, scaling, and tubing. - Two samples of produced water (20 L each) were collected from a production facility in Kuwait in November 2021 and in January 2023. Three samples of scaling (200 g each, named here scalings 1 - 3) were collected from a production tubing assembly that carried oil and produced water as well as small amounts of gas at the same production facility in January and in February 2023 (Figure 2). For this, production piping (diameter: 3.5 inches) was dissembled (Figure 3, Figure 4), and the scaling was flushed with diesel. Before collection, the diesel was evaporated from the scaling. In addition, the authors received a completely de-oiled scaling sample from the same production site (here, named scaling 4). A short section of a used tubing piece at end-of-life was obtained from the production site. Pieces of the tubing were mechanically cut to mount them for the necessary analytical analyses.

Determination of the composition of the scaling and the tubing. – The scaling which included some remaining moisture contained both hydrocarbons and metal salts. To determine the moisture and volatile content, the scaling samples were heated to 120˚C for 24 h in a Carbolite oven (ELF 11-6) with a weighed sample of each product placed in a crucible (79C-00, Waldenwanger, Berlin). The moisture/volatile contents are expressed in w% within the text. A triplicate of experiments was carried out and the values are given as average values ± standard deviation. To determine the metal content, the scaling was heated at 600˚C for 3 h with a weighed sample of each product placed in a crucible (79C-00, Waldenwanger, Berlin). The ash contents are expressed in w% within the text. A triplicate of experiments was carried out and the values are given as average values ± standard deviation. As metal salts can potentially react under the ashing conditions, especially iron sulfide to iron oxide and low valency iron species to iron (III) oxide, an extraction of the raw scalings with toluene was performed. For this, scalings (ca. 10 g) were weighed into a round-bottom flask, toluene (70 mL) was added and the resulting suspension was stirred at rt for 4 h. The toluene solution was decanted, and the process was twice repeated with 70 mL toluene, each. Finally, the extracted scalings were filtered off and dried in an EcoCell drying cabinet at 37˚C for 12 h. A triplicate of experiments was carried out and the values are given as average values ± standard deviation. The cooled ash and toluene extracted samples were subjected to wavelength dispersive X-ray fluorescence spectroscopy (WD-XRF), powder X-ray diffraction (P-XRD), scanning electron microscopy coupled with energy dispersive X-ray spectroscopy (SEM-EDS), X-ray photoelectron spectroscopy (XPS) as well as to FT-IR spectroscopy (FT-IR).

Wave-length dispersive X-ray fluorescence (WD-XRF) -The elemental composition of each sample was analyzed using a wavelength dispersive X-ray fluorescence (WD-XRF) instrument (Rigaku ZSX Primus IV, Japan) equipped with a Rh X-ray tube and controlled by ZSX guidance software. The sample was placed on a specified sample holder cup and analyzed directly.

Powder X-Ray diffraction analysis (P-XRD)—The powder X-ray diffraction analysis was recorded on a Rigaku Miniflex XRD, Japan, where the diffraction patterns of the samples were recorded with a Bragg’s angle 2θ from 5˚ to 90˚ using Cu Kα radiation (λ = 1.54 Å) at voltage 45 kV and 40 mA current. All the analyses were performed at ambient temperature.

Scanning electron microscopy coupled with energy dispersive X-ray spectroscopy (SEM-EDS)—SEM images were captured with a Thermo Quattro S instrument at high vacuum mode (6 × 10⁻⁴ Pa), with an accelerating voltage of 30kV and using an Everhart-Thornley SE detector (ETD). The EDS analysis was performed with an energy dispersive X-ray detector attached to the SEM.

X-ray photoelectron spectroscopy (XPS)—A photoelectron spectrometer Nexsa G2 (Thermo Scientific, UK) was employed in the conducted studies to examine the elemental composition and chemical state of the samples. The instrument utilized monochromatized Al-Kα radiation (1486.6 eV) with a spot size of 400 μm and employed a flood gun for static charge compensation. Survey spectra were acquired with a pass energy of 200 eV, while high-resolution scans were conducted at 50 eV, all under an ultra-high vacuum environment of approximately 10⁻⁹ mbar.

Fourier Transform infrared spectroscopy (FT-IR)—The FT-IR spectra were obtained using a Perkin Elmer Spectrum Two Spectrometer and the data were processed using Spectrum IR software. For sample preparation, IR grade KBr (Aldrich) was mixed with the sample and the transmittance of the ensuing KBr sample pellet was recorded in the range 4000 - 500 cm⁻¹ and processed after 32 scans.

The character of the produced water from the sour oil field causes extensive corrosion of the subsurface equipment including tubing. The corrosion is especially induced by the high chloride concentration. Of interest is the relative paucity of sulphur containing species measured in the collected produced water. Usually, H₂S distributes itself mainly over the gas fraction (Table 7), and a small amount (ca. 1v% of the total) remains in the hydrocarbon fraction. Although H₂S is soluble in water (3.98 g/L at 20˚C), its solubility decreases with increasing salinity of the water (about 2.5 g/L in seawater with TS of 35.000) [29] . Before the investigated produced water had been collected, flashing and stripping occurred, which would remove most of the H₂S present. However, the produced water was not specifically deaerated. Thus, the concentrations of sulphur species in the produced water do not reflect the concentrations of the sulphur containing molecules, including H₂S, that the tubing is exposed to. That these combined concentrations, especially when taking into account the H₂S concentration in the gas fraction (Table 7), are much higher, can be seen in the relatively high sulfur concentrations of the corrosion patina on the inner tubing.

The nature of the production fluid leads to extensive corrosion of the subsurface tubing with scalings blocking the production tubing. Scaling samples 1 - 4 collected from different blockages of the tubing have been found to be very heterogeneous (Table 4). In operating systems with a larger number of wells grouped around a central collecting unit the composition of the scalings will vary with which wells are operating at that moment and with what capacity. From XRD, it can be noted that the iron salts that make up part of the scaling are magnetite, hematite, greigite, siderite and goethite. Similar oxidation products from corrosion had been reported from the tubing from another Kuwaiti oilfield [8] .

SEM images of debris stemming from the tubing that is part of the scaling show geometrically aligned rows of pits, which have partially progressed to the state of perforations. The pits and the perforations are not necessarily concentric, but have oblong shapes with the main axes aligned, most likely influenced by residual stress in the material.

Typically, tubing replacement happens every six months at a cost of $12.9 per tubing meter. This replacement contributes to the poor financial return of mature sour oilfields. In regard to mature oilfields, it must be kept in mind that a strategy needs to be in place to handle large amounts of produced water, and

that in case of re-injection into the production well or in case of a general subsurface disposal, an adequate treatment of such large amounts of water exists. Currently, both options, disposal, and injection, exist, although presently the produced water is pumped into a disposal well.

This contribution examined the impact of produced water from a sour oilfield in South Kuwait on the production tubing in the subsurface. It was noted that the heavily saline produced water leads to significant corrosion of the steel tubing discharging debris into the tubing. The debris can lodge in the tubing and was handled in this contribution as part of the scaling. The debris exhibited geometrically aligned corrosion pits, some of which had progressed to perforations. The used tubing itself showed a patina made up of iron oxides (such as hematite), iron carbonate (siderite) as well as iron sulfides (such as greigite). X-ray photoelectron spectra (XPS) of scaling samples, for the most part debris, showed that iron [Fe(0)] was still very much in evidence on the surface, along with Fe(II) and (Fe(III) species. Calcium salt deposits on the inner tubing at end of life made up of 2.6 w% the scaling directly adhering to the tubing. Silica deposits were also in evidence. The debris showed 0.75 - 9.3 w% Ca and 3.4 - 8.5 w% Si content.

In this type of operation, the produced water is either reinjected into the production well or is disposed in a disposal well. Especially in case of reinjection the produced water needs to be treated. The authors have not yet examined the above-ground tubing and the tubing used for reinjection or disposal. Nevertheless, it is recommended that produced water is treated at an early stage to minimize damage to above-ground and reinjection tubing. Lastly, produced water may represent a non-conventional water resource in addition to a source of industrial grade salt, if treated adequately.

The authors gratefully acknowledge Javad B.M. Parambath, Advanced Materials Research Laboratory, University of Sharjah for XPS analysis and interpretation of the results. The authors thank Wajeeh Kittaneh, Department of Geosciences, UAEU, for the use of the stereoscope (Model SZ2-ILST). Also, the authors are grateful to Afra Alblooshi, Department of Chemistry, UAEU, for help in operating the SEM-EDS instrumentation

The authors declare no conflicts of interest regarding the publication of this paper.

Al Salem, F., Poulose, V., Kawamura, K., Nakamura, A., Saibi, H. and Thiemann, T. (2023) Effects of the Produced Water from a Sour Oilfield in South Kuwait on the Production Tubing. Journal of Water Resource and Protection, 15, 358-375. https://doi.org/10.4236/jwarp.2023.157021