Optimization of oil recovery using heated low salinity water (HLSW) in the horizontal sand pack column during water flooding: radiotracer intervention

Water-flooding is a prevalent technique for secondary oil recovery that is capable to increase oil recovery by up to 45%. Due to its accessibility, affordability, and simplicity, water flooding is the most frequently used secondary oil recovery technique. However, the efficiency of conventional water flooding is limited by capillary forces which holds the residual crude oil in pore structures. Moreover, during water flooding operation, the injected fluid does not penetrate sufficiently due to low permeability or presence of channelling inside the sandstone reservoir. Therefore, in this study, thermal recovery was introduced to 100ppm (0.1g/L) low salinity water (LSW) in order to investigate the percentage of oil recovery during water flooding. A horizontal sand pack column was used for water flooding experiment and temperature and injections rate are the parameters to be optimised with respect to percentage of oil recovery. The column was initially packed with 125micrometer sieved sand to ensure constant homogeneity is achieved. Initially, formation water which comprised with NaCl, CaCl2 and MgSO4 was introduced inside the column together with light oil and was aged overnight. TAPIS oil with viscosity of 0.001382 Pa.s (1.382cP) was used as light-oil in this study. The injection rate was set to 1 mL/min, 2 mL/min, and 3 mL/min respectively whereas, the temperature of the LSW was heated at 50 °C, 70 °C, and 90 °C subsequently. Heated water was supplied continuously throughout the sand pack which was heated using water blanket to retain heat at 70°C (reservoir temperature). In conclusion, 3 mL/min and 70°C were identified as optimum parameters and the oil recovery experiment was proceeded with liquid radiotracer using Technecium-99m (Tc-99m) intervention and resulted with 73% of yield. Nevertheless, the radiotracer intervention only provides the Residence Time Distribution (RTD) model which describe the behaviour of the sand pack during water flooding process.


Introduction
Water flooding is one of the most utilized Enhanced Oil Recovery program in order to extract oil in place from matured reservoir.Recently, low high salinity water (LSW) has attracted attention of the stake holders and numerous researches have been carried out to optimize the yield of oil from the oil 1308 (2024) 012002 IOP Publishing doi:10.1088/1757-899X/1308/1/012002 2 reservoir.Previous authors claimed that LSW is able to extract additional of 5-38% from the reservoir unlike high salinity water that cannot tolerate with the clay particles attached to the pore walls due to wettability issues thus hinder the oil recovery [1].It is believed that the kaolinite from the pore walls, which is the clay particles will detach easily by the presence of LSW.The oil that is clinging together with moveable kaolinite clay at the pore walls will be easily swept thus, enhances the recovery of oil.According to [2], the recovery of oil is significantly enhanced when the salinity of injecting water for water flooding activity is lower than connate water.Moreover, LSW injection is most preferred because of its cost effectiveness due to the availability of water and cheaper than other approach plus the oil displacement is slightly easier due to its excellent maneuverability in oil reservoir.Seawater, fresh water and injecting water from other reservoirs or produced water are the most common water sources for water flooding operations [3].However, water flooding is best applied to light crude oil reservoir.The crude oil that we used for this study is from TAPIS reservoir which has viscosity of 0.001382 Pa.s (1.382cP) and considered as light oil.Light crude oil is defined as having low density and viscosity and move freely in room temperature.
Malaysia reservoir is basically fractured basement rock reservoir located offshore.The structures are characterized by a fracture system of very high heterogeneity that leads to anisotropic flow [5].Thus, radiotracer technology is proposed due to its ability to permeate through any voids and soil stratification.This is due to the fact that liquid tracer can squeeze through any soil openings that leads to information of the reservoir as well as displacing the residual oil.Basically, analysis of residence time distribution (RTD) is carried out in order to determine model that fits the reservoir well before thorough interpretation of the respective reservoir can be made.Hence, the primary objective of this study is to optimize parameter in enhancing oil recovery for sandstone reservoir.Thus, heated low salinity water (HLSW) is proposed to substitute conventional LSW water flooding in order to investigate the effectiveness of thermal expansion during water-flooding activity.In this study, the sand column is heated at 70 o C throughout the experiment to mimic the temperature of actual reservoir.Moreover, this study aims to investigate the hydrodynamics behavior inside the porous domain by providing the analysis of residence time distribution (RTD) models using radiotracer intervention.

Water flooding set-up: optimization parameters
Figure 1 shows the schematic diagram of overall experiment for optimization parameters.The experiment was started with an empty column compacted with 125 micron -sized sand.The fully loaded column was weighed to obtain the mass of sand in the rig.Next, formation water was prepared as shown in Table 1 and injected inside the column using peristaltic pump followed by 100 ml of kerosene subsequently and left for a day.Brine solution was prepared inside a beaker and heated up to designated temperature.Both beaker and column were wrapped with aluminum foil to maintain the temperature.Finally, then brine solution was injected inside the column using peristaltic pump at designated flow rate for 6 hours.The injection rate was set to 1 mL/min, 2 mL/min, and 3 mL/min respectively whereas, the temperature of the LSW was heated at 50 °C, 70 °C, and 90 °C accordingly.

Radiotracer experiments
Radiotracer experiment was carried out on the elevated core flood rig after optimum parameters (temperature and flow rate) were obtained from 2.1.The radioactive experiment was conducted once in order to reduce the radiation exposure to the operator as well as for safety purpose.Similar procedure was adopted as 2.1 except in this experiment 6 sodium iodide (Tl) scintillation detectors were arranged externally along the column.All detectors were connected to Data Acquisition System (DAS) which was connected to laptop for LSW-tracer monitoring.Technetium-99m, a gamma emitter water tracer with 0.140MeV was used for water flooding activity.About 5.37mCi/3ml of Tc-99m was injected inside the column while LSW was running continuously.

Optimization of water flooding parameters
In this study, the injection rate was set to 1 mL/min, 2 mL/min, and 3 mL/min respectively whereas, the temperature of the LSW was varied at 50 °C, 70 °C, and 90 °C accordingly.According to Afiq et al. [3], when injection rate increases, the capillary force is increased that leads to high sweep efficiency and improved oil displacement.The existence of fingering is also avoided by selecting the lower flow rate for water flooding injection.Besides, the low viscosity light oil is used which reduces the presence of viscous fingering.Similarly, as temperature increases, the yield is significantly improved.This is due to the fact that hot water injection is a thermal recovery technique which reduces significantly the viscosity of oil.Thus, by thinning the oil viscosity it provides the driving mechanism for the hot water to push the residual oil freely.Moreover, the Low Salinity Water provides additional mechanism to detach the oil from the sand wall easily hence improve oil displacement and achieve greater oil recovery result.This is because the interaction of ions from the water phase with carboxylic acid from the oil phase leads to changes in interfacial tension (IFT) of the fluids.From the compiled findings [4], whereby the salinity under investigations were 15000-100000ppm, it was concluded that IFT increases with increasing salinity.Thus, since 100ppm is used for the study, the IFT between LSW-oil is decreased and the recovery of oil is significantly increased as compared to previous authors as shown in Table 2.In conclusion, the oil recovery is proportional to temperature and flow rate in the range of temperature and flow rate used in this study as shown in Figure 3

Radiotracer experiments: qualitative analysis
Figure 4 shows the radioactive signals acquired by each detector once the water flooding commences.The continuous water injection carries Tc-99m which emits gamma rays and the scintillation detector from individual detector is picking up the signal.D1 portrays highest and sharp peak which implies the radioactive material has highest concentration and sufficient strength prior entering the column.
Similarly, the peak is still significant at D2 because the mixing of HLSW-tracer is about to start inside the sand column and the pressure is relatively low.As the HLSW-tracer moves from right to the left of column, the resistance inside the column is very dominant.The pressure is increasing since the HLSW-tracer has to loose itself from the packed sand, formation water and residual oil respectively in order to permeate and distribute homogenously inside the packed column and finally towards the end of the column.Therefore, the distribution of the curve from D2 until D5 is gradually broaden and flattened.Moreover, it can be inferred that D1 is producing high level of counts compared to all detection sensors because there is no resistance experience by D1 since the detection is carried out outside of the sand column.Figure 1 indicates that all detectors are able to pick up radioactive signals which are due to fluid flow from water-flooding activity of HLSW-tracer inside the column.

Radiotracer experiments: quantitative analysis
In order to determine the process anomalies inside the sand pack during water flooding, analysis of residence time distribution (RTD) is carried out.Basically, assigned RTD model enables to provide the deviant of the respective process from ideal condition.The International Atomic Energy Agency (IAEA) has recommended six RTD models to be fitted with experimental RTD derived from radiotracer experiment [9].Each model has its own parameters that is optimized in order to select the best fitting curve.Hence, the sum of square errors (SSE) is used as means to justify the optimum RTD model to tracer response experiment.The least value of SSE, the minimum deviation of between the experimental data and the model function, is the benchmark for model selection among the recommended RTDs [10].The mathematical expression for RTD is shown in Equation (1).The E(t) function is the RTD where E(t) dt is the fraction of the flow, measured at the exit, that is in the system between times t and (t +dt) [9]: where C(t) is the concentration of radiotracer monitored by NaI scintillation detectors in counts per second (cps) as numerator and denominator is the area under the curve of plotted C(t). Figure 5 shows the data from Detector 6 which is the response of the tracer concentration profile at the outlet of the sand pack and the concentration measurements are recorded using data acquisition system (DAS) as shown in Figure 2.This data is rich in information that provides the information of process behavior during the water flooding process.The data is then transformed using Eq.[1] and the RTD models against experimental data are plotted using IAEA RTD Software as shown in Figure 6.From observation, the perfect mixer in series with exchange is the most representative model than other models because the SSE is 0.364 x 10 -9 , the least value compared to other models as shown in Table 1.The perfect mixers in series with exchange consists of several parameters which are τ, J, Tm and K where all the variables are defined as follows: In this model, the mechanism of flow can be described in Figure 7.The flow is basically distributed into two distinctive volumes; V1 which is total volume of the primary stream and V2 total volume of secondary stream respectively.The secondary flow comprises with volume of exchange cells with interconnected αQ flow rate.As shown in Table 3, the number of perfect mixers of the respective column is 4 with mean residence time of 1824s for primary zone and 4800s for secondary zone respectively.The high value of Tm can be translated to the migration of ions from primary to secondary flow.K value is 3.3 which indicates the volume of primary zone is superior three times than secondary zone and thus, reflects the occurrence of an exchange flow rate αQ between the main flow and secondary flow whereas Q is the volumetric flow rate of the primary stream from one volume to another.The presence of liquid ions in formation water (Na + , Cl -, Ca 2+ , Mg 2+ , SO4 -), brine (NaCl) and tracer (Tc O 4 −¿ ¿ ) contribute to the exchange of ions between primary stream and secondary stream.
Moreover, the interaction of HLSW-tracer has reduced the IFT and perhaps altered the wettability of sand towards water-wet, thus increase the oil production.Wettability alteration is required to ensure the wetting state of rock is dominant towards water wet since wettability alteration is an effective approach to enhanced oil recovery.(2)

Conclusions
In this study, two main analysis were carried out successfully which were optimization of residual oil from the heated low salinity water (HLSW) and determination of RTD model that represents the response of the tracer concentration profile at the outlet of the sand pack.Basically, the chosen RTD model describes the behavior or characteristics of the respective column.In this study, the optimum parameters for oil recovery is 3 mL/min and 70 o C with yield of 73% of oil.This oil recovery is the highest in comparison to oil recovery from previous authors.Moreover, the perfect mixers with exchange model describes the sand column well due to the observed value of SSE is the minimum among all recommended models.The model defines then presence of four perfect mixers and has two distinctive volumes which are primary and secondary volumes.Both volumes are connected with an exchange flow rate denoted as αQ and the presence of ions are basically from formation water, brine and tracer respectively.

Figure 1 .
Figure 1.Schematic diagram of experimental set up.

Figure 4 .
Figure 4. Radiotracer experiments: Fluid flow of HLSW-tracer inside the sand pack column.

Table 1 .
Composition of formation water and low salinity water.

Table 2 .
Oil recovery from various salinity concentration.

Table 3 .
Sum of Squares Sum of Squares Error