Received 22 June 2016; accepted 6 August 2016; published 10 August 2016
Oil and gas production and associated processing operations often result in the accumulation of Naturally-Oc- Curring Radioactive Material (“NORM” or “TENORM” if technologically enhanced) at elevated concentrations in by-product waste streams. The ultimate sources of most of the radioactivity are from daughter products of uranium (238U, 235U) and thorium (232Th) that are naturally present in subsurface formations from which oil and gas are produced. NORM, in the form of scale, has been known for many years to occur in some oil and gas pipelines. Deep formation waters, often with high total dissolved solids, tend to be enriched in natural radium isotopes. When these fluids are brought to the surface, CO2 may escape, resulting in a pH rise with the solubility products of many sulfate and carbonate species being exceeded. The result is often precipitation of scale deposits rich in the divalent alkaline earths Ca2+, Ba2+, Sr2+, and Ra2+ (Wilson and Scott, 1992; Mackay and Heriot- Watt, 2003; Efendiyev, 1953)  -  . Radium, and its decay products, can occasionally be concentrated enough to be of human health concern because of their gamma radiation (Bernhardt et al., 1996; Smith et al., 1996)   . It is estimated that each year in the oil/gas industry between 300,000 and 1,000,000 metric tons of NORM scale are produced (Tomson et al., 2003)  . The primary radionuclides of concern in NORM wastes are 226Ra (238U series), and 228Ra (232Th series) as well as their respective daughters.
The production waste streams most likely to be contaminated by elevated radium concentrations include produced water, scale, and sludge. Spills or intentional releases of these waste streams to the ground can result in NORM-contaminated soils that may need to be disposed. Since radium is fairly soluble in saline waters, it can follow the produced water stream and accumulate in formation water storage ponds rather than all being retained in scale deposits. Thus, dissolved radium in formation waters either remains in solution in the produced water, precipitates out in scales or sludges, or is removed from the produced water during storage (e.g., precipitation, adsorption onto sediment components). Conditions that effect radium solubility and precipitation include water chemistry (primarily salinity), temperature, and pressure (Jens and Sebastian, 2002)  .
NORM contamination of scale and sludge can occur when dissolved radium coprecipitates with other alkaline earth elements such as barium, strontium, or calcium. In the case of scale, the radium coprecipitates primarily with barium, to form hard insoluble sulfate deposits (barite) (Doerner and Hoskins, 1925)  . Such scale typically forms on the inside of piping, filters, injection wellhead equipment, and other water handling equipment, but also can form as a coating on produced sand grains. In the case of sludge, radium can be present in several forms. It can coprecipitate with silicates and carbonates that form in the sludge, or it can be present as particles of barium sulfate scale that become incorporated into the sludge. NORM-contaminated sludge can accumulate inside piping, separators, heaters and other types of treatment facilities, storage tanks, and any other equipment where produced water is handled.
Formation waters that are sent to disposal ponds may consist of accumulated heavy hydrocarbons, paraffin, inorganic solids, and heavy emulsions. We report here results of chemical and radiochemical measurements from formation waters separated from oil pumped from approximately 700 m below ground near
2. Study Site and Experimental
We have chosen a study location that represents one of the original sites for oil production in
The petroleum began to be extracted at this site from the beginning of the last century (1907 with other sites in this area developing since 1912) (Alizadeh and Akhmedov, 1966)  .
Figure 1. Map of Absheron peninsula showing site of oil field lake. (This site is located in the southeast portion of the Absheron Peninsula, near the capital city of Azerbaijan Republic-Baku. The petroleum began to be extracted at this site from the beginning of the last century-1907, with other sites in this area developing since 1912).
In recent years, the level of the deposits has been falling in the Surachany region requiring deeper drilling with lower grade deposits (higher percentage of formation water). At this site, as well as on other areas of the
We collected samples of lake water, lake sediments, and formation waters from 7 different oil wells that surround the radium lake. All water samples were analyzed for major and minor cations (Na, K, Ca, Mg, Sr, and Ba) by atomic absorption spectrophotometry and anions (Cl, HCO3, SO4, Br, and I) by standard wet chemical techniques (B. Suleymanov et al., 2008)  . We also measured the pH and total dissolved solids (TDS) for each water sample. Both water and sediment samples were analyzed for 226Ra, 228Ra, and other radionuclides via gamma-spectrometry using a
3. Results and Discussion
3.1. General Situation
There are 86 wells operating in the immediate area around the formation water lake. These wells are pumping from an average depth of about 700 m and are recovering a very low grade oil deposit with most pumped fluids having oil content considerably less than 2%. After separation of the oil the formation water is released to the “lake”. These formation waters are all very high in total dissolved solids (>70 to 160 g/L). While the in situ temperatures are approximately 60˚C, the water discharged to the lake is about 40˚C after separation. Approximately 3000 m3/day accumulates in this manner. When the level of the lake becomes too high, water is pumped into the Caspian Sea. Based on an estimated volume of the lake, the average residence time of the water is on the order of only a few months.
3.2. Well and
Table 1. Concentration of cations and anions in well (W) and lake (L) waters. All samples filtered through a 0.45 m filter except for the carbonate analyses.
Table 2. Radionuclides in well (W) and lake (L) waters.
They are all very high in TDS, mainly a sodium chloride brine with secondary calcium bicarbonate. All are in a pH range of 6 - 7. The waters are also very high in radium, with 226Ra activities as high as 2.4 Bq/L and 228Ra activities up to 5 Bq/L. Since “average” groundwater activities are typically around 0.015 Bq/L (Eisenbud and Gesell, 1997)  , these waters are elevated by over two orders of magnitude. In general, the radium isotopes follow each other (Figure 2) with an average activity ratio of 2.0 ± 0.4.
Both 226Ra and 228Ra tend to be highest in the waters with the highest TDS (Figure 3). The lake water displays an intermediate character in this respect as it does with many of the other characteristics. This may be a result of the relatively low residence time in the lake as this is just temporary storage until the water is pumped into the
While it is not clear what is controlling the radium content in these waters, we note that there is a relatively strong inverse relationship to bicarbonate (Figure 4). While there is also an inverse relationship to sulfate, the bicarbonate dependency appears stronger. There is actually a somewhat positive relationship between radium and barium, which is surprising as barite (BaSO4) is often thought to control radium solubility.
It may be that there were fine particles of suspended barite in the water samples (samples for gamma spectrometry were filtered through a paper filter to remove oil residues, etc., but fine particles could have passed through). We entered the chemical data for the lake water (sample 8 L) into the thermodynamic program MINEQL and it predicted that barium sulfate and calcium carbonate would precipitate. The program calculated that 99.6% of the Ba would precipitate out as BaSO4, and 36.6% of the Ca as CaCO3. Our chemical results show some Ba is in solution (0.4 m filtered) so it seems unlikely that such a large fraction precipitated. The radium- bicarbonate relationship (Figure 4), however, suggests that carbonate precipitation in these waters is significant and may be an important control on the radium concentration.
Figure 2. Ra-228 versus 226Ra in well waters and lake water. The open diamonds are well waters and the closed symbol represents the lake water.
Figure 3. Radium isotopes versus total dissolved solids.
We analyzed radium isotopes from 2-cm thick layers from a sediment core collected in the central portion of the lake in June 2006. All of these samples were analyzed in sealed (~100 cm3) plastic containers using the same gamma spectrometer as the lake waters. The results (Table 3) show that the sediments are somewhat (approximately a factor of 2) enriched in radium compared to typical soils and sediments (Eisenbud and Gesell, 1997; Huck et al., 1989)   but not nearly elevated to the same extent as the lake waters.
In addition, the isotopic composition of the sediments is very different from the formation waters with an average activity ratio of 0.89 ± 0.13 (formation waters AR = 2.0 ± 0.4). In fact, the ARs in the youngest sediments are still lower, around 0.60. It thus appears that if radium is precipitating and being added to these sediments, the additions must be relatively minor.
When one examines the 228/228Ra AR down core, there is a trend of increasing values (Figure 5) until a fairly steady-state level is obtained.
Figure 4. Radium isotopes versus bicarbonate concentration. The outlier at 70 mmol/L HCO3 may be a result of carbonate particles in suspension that consumed acid during the titration. Samples for carbonate analyses were run unfiltered.
Table 3. Natural radionuclides in sediment layers from a core in the radium lake.
Figure 5. Distribution of the 228Ra/226Ra activity ratio (AR) with depth in a sediment core from radium lake. The dashed line represents ingrowth assuming that 228Ra was below equilibrium with 232Th in the young sediments near the top and reached equilibrium at a depth around 10 cm.
If this trend is related to in growth of 228Ra into its parent 232Th, the 5.7-year half-life of 228Ra would require about 25 years to reach the quasi-steady-state level at about 10 cm depth (accumulation rate of ~ 0.4 cm/yr). Unfortunately, the upper levels of the core do not have any excess 210Pb, so we cannot confirm that age estimate using 210Pb chronology.
We did verify that the upper two layers in the core (up to 4 cm) contained fallout 137Cs so at least these layers are no older than about 40 yrs.
The investigated water samples are very high in radium, with 226Ra activities as high as 2.4 Bq/L and 228Ra activities up to 5 Bq/L. Both 226Ra and 228Ra tend to be the highest in the waters with the highest TDS. There is actually a somewhat positive relationship between radium and barium, which is surprising as barite (BaSO4) is often thought to control radium solubility. There is a relatively strong inverse relationship to bicarbonate and sulfate, and the bicarbonate dependency appears stronger. The radium-bicarbonate relationship, however, suggests that carbonate precipitation in formation waters is significant and may be an important control on the radium concentration.
When one examines the 228/228Ra AR down core, there is a trend of increasing values until a fairly steady- state level is obtained. If this trend is related to ingrowth of 228Ra into its parent 232Th, the 5.7-year half-life of 228Ra would require about 25 years to reach the quasi-steady-state level at about 10 cm depth (accumulation rate of ~ 0.4 cm/yr). We did verify that the upper two layers in the core (up to 4 cm) contained fallout 137Cs so at least these layers are no older than about 40 yrs.
Investigation of radium isotopes (226Ra and 228Ra), 137Cs and useful parameters of formation water give important information about history and origin of NORM contamination. The radium-bicarbonate relationship may be an important control on the radium concentration, without expensive radionuclide analyses.
The authors’ sincere thanks are due to Dr Burnett W. C., from Florida State University, USA for his essential advice provided throughout this work.