IODP Proceedings    Volume contents     Search


Inorganic geochemistry

Drilling mud

Drilling mud contamination presents a potential challenge when collecting, processing, and interpreting the composition of interstitial water from sediment cores obtained through riser drilling (Wenger et al., 2004). In order to address this problem, we first review the aspects of the drilling mud composition that provide a context for discussing variations in the interstitial water composition at depth. Throughout drilling of Hole C0020A, the formulation of the drilling mud was adjusted in response to changes in drilling procedures and hole conditions that occurred with depth. In order to provide a working depth reference of the drilling mud for comparison with the interstitial water of the sediment cores, we present the compositional changes in mud samples obtained from the active circulation tank in the context of the lithologic boundaries that were being drilled at the time of their collection. These units (see “Lithostratigraphy”) are as follows: Unit I (647–1256.5 m MSF), Unit II (1256.5–1826.5 m MSF), Unit III (1826.5–2046.5 m MSF), and Unit IV (2046.5–2466.0 m MSF).

A total of eight drilling mud samples were sampled from the mud tank (LMT samples) in order to assess potential contamination during the course of drilling (Table T12). The first mud sample (337-C0020A-2-LMT) was collected 11 days before the commencement of drilling. Shortly after the initiation of the drilling in Unit I, a second sample was taken from the mud tank (Sample 23-LMT). Two mud water fluid samples were collected from core liner (LMW samples) for analysis from Unit II (Samples 337-C0020A-120-LMW and 189-LMW), one from Unit III (Sample 257-LMW), and two from Unit IV (Samples 280-LMW and 373-LMW). A final mud sample (392-LMT) was taken from the active mud tank immediately after termination of drilling operations and may be considered representative of the mud used in the final wiper trip of the borehole before logging and formation water sampling. The water was separated from the mud (see “Inorganic geochemistry” in the “Methods” chapter [Expedition 337 Scientists, 2013b]) and analyzed. A list of miscellaneous materials collected for analysis is found in MISC_MATERIAL_SAMPLE_LIST.XLS in GEOCHEM in “Supplementary material.”

The shipboard pmH of the mud fluid was not recorded, as the samples had to be diluted 1:10 to carry out the mud water separation. The mud logger’s report, however, indicates that during the course of drilling, the mud mix pmH values ranged from 9.3 to 11.9. The mud logs also reveal a number of significant modifications to the drilling mud that account for the recorded changes in mud water chemistry. Initially, and throughout the course of drilling in Unit I, KCl was combined with the mud, giving a mud density of 1.04–1.08 g/mL. During drilling in Unit II, NaCl was added to the mud along with NaOH, maintaining the mud density at 1.08 g/mL. In Unit III, ASTEX-S sulfonated asphalt sodium salt (SAS) was added to enhance hole stability as well as to enhance the lubrication characteristics of the drilling mud. The NaCl + KCl + SAS + NaOH mixture was continued throughout Units III and IV, and mud density was maintained at 1.10 g/mL until reaching the final depth of 2466 mbsf. Filtered mud water changed from colorless in Units I and II to a dark amber color in Units III and IV. Significant changes were made to the mud formulation when drilling through Unit III, and the change in coloration may be the result of the addition of ASTEX-S SAS to the mud or it could be the result of colloidal-sized particles that were released from the sediment during riser drilling of the coal-bearing intervals located in the lowermost portions of Unit II, as well as Units III and IV.

Total alkalinity of the mud water increased from 39.7 to 49.9 mM across Units I and II. Alkalinity then increased abruptly to 77.7 mM in Unit III, followed by a slight drop to 70.03 mM at the bottom of Unit IV. The abrupt change between Units II and III was a result of a change in the drilling mud formulation that involved the addition of NaOH. Chloride concentrations in the mud ranged from 970 to 1131 mM across Units I and II and then abruptly increased to 1548 mM in Unit III. Chloride values decreased somewhat in Unit IV, ranging from 1367 to 1394 mM, which is still roughly twice that of average seawater (559 mM). Potassium concentrations in the drilling mud were also significantly greater than seawater (10.5 mM) because of the initial addition of KCl to the drilling mud (Fig. F24A). A rough mass balance of the ions analyzed in the initial mud (Sample 2-LMT) shows that roughly ~600 mmol/L of KCl was added to seawater to bring the mud up to the measured chloride values, and then an additional ~200 mmol/L of potassium as KOH was used to adjust the pmH. Potassium concentrations throughout the course of drilling ranged from 756 to 662 mM, with somewhat lower values in Unit IV, where the mud was reformulated to contain NaCl and NaOH in addition to KCl and KOH.

The sulfate concentration of Sample 2-LMT (27.95 mM) prior to drilling was similar to that of seawater (28.9 mM) (Fig. F24B). The sulfate content of mud water was adjusted to lower values during drilling in Units I and II (20.5–25.7 mM); however, a sharp increase in Unit III (56.0 mM) was followed by elevated values in the mud used throughout drilling in Unit IV (9.4–51.0 mM). These elevated values were a consequence of the addition of ASTEX-S SAS to the drilling fluid, which includes Na2SO4 as a major component. Sodium concentrations also paralleled those of chloride and sulfate, with sodium generally close to or exceeding seawater (480 mM). Sodium concentrations of the mud used in Units I and II ranged from 457 to 624 mM, increased to 1147 mM in Unit III, and decreased slightly to 1014 to 1020 mM in Unit IV. Salinity measurements showed similar trends ranging from 98‰ to 103‰ in Units I and II, increasing to 139‰ in Unit III, and decreasing again with values between 126‰ and 127‰ across Unit IV (Fig. F24C).

Because bromide was not included in the mud water additives, concentrations remained similar to those of the seawater (0.85 mM) that was mixed with the drilling mud. The measured bromide concentrations ranged from 0.68 to 0.94 mM across Units I–IV. A number of ions potentially present in deep interstitial water were notably below instrumental detection limits in the drilling mud samples, including NH4+, NO3, and ΣPO43–. Magnesium and calcium concentrations are consistently less than seawater values (54.0 and 10.6 mM, respectively). Prior to drilling, the recorded Mg2+ concentration was 28.2 mM and gradually decreased with depth to 2.23 mM at the bottom of Unit IV. Likewise, Ca2+ started at 5.96 mM before drilling and decreased to 4.05 mM at the bottom of Unit IV.

Dissolved silica increased significantly during drilling. The total dissolved silica content of the mud started at 0.12 mM and increased to only 0.23 mM in Unit I. However, concentrations increased to between 2.31 and 2.15 mM in Unit II, 4.42 mM in Unit III, and 5.28–7.20 mM in Unit IV. The increase may have occurred because of the addition of ASTEX-S SAS, which contains crystalline silica. It was also probably caused, in part, by the increased solubility of silica at high pH (Eikenberg, 1990; Southwick, 1985). Dissolved barium should be negligible in the mud because of the presence of sulfate from seawater (e.g., Torres et al., 1996) and from the added ASTEX-S SAS. Nonetheless, barium is found in the drilling mud samples in concentrations ranging from 5.1 to 42.6 µM across Units I–IV. The initial mud concentrations are the lowest, and most concentrations are between 9 and 15 µM.


Coring was not performed during drilling of Unit I; however, drilling cuttings were collected from the shale shaker, which separates sediments before returning the mud to the circulation tanks. Shipboard analyses of cuttings water samples continued through Unit II (Table T13) in order to provide a comparison between the cuttings water and interstitial water from sediment cores. A total of eight cuttings samples were processed and analyzed: three from Unit I (Samples 337-C0020A-56-SMW, 66-SMW, and 81-SMW) and five from Unit II (114-SMW, 128-SMW, 187-SMW, 197-SMW, and 213-SMW). The first sample (56-SMW) was done in replicate, and sediment fragments in the first portion were cleaned prior to further processing by manually wiping the adhering drilling fluid from individual cuttings pieces with KimWipes, whereas the other portion was processed in bulk without cleaning off drilling fluid. The second sample (66-SMW) was manually cleaned in a similar fashion, whereas all other samples were done in bulk because of the difficulty associated with cleaning small cuttings fragments. Within Unit II, the fluid from one sample (197-SMW) was collected using a Rhizon sampler (Seeberg-Elverfeldt et al., 2005) in order to determine whether there would be a significant difference between this procedure and that of the squeezed samples (see Fig. F24).

The results from the cuttings water bore many similarities to the drilling water, including elevated amounts of potassium ranging from 621 to 781 mM (Fig. F24A), high sulfate concentrations of 18.6–23.8 mM (Fig. F24B), increased salinity values of 70‰–103‰ (Fig. F24C), and high chloride concentrations of 1129–1515 mM, that are all indicative of drilling fluid contamination. Nonetheless, some significant differences were found between the water squeezed from the cuttings and that of the drilling mud water. In Unit I, ammonium and phosphate were present in the cuttings water but were below detection in the drilling mud fluid. Ammonium ranged from 11.0 to 3.7 mM in Unit I and decreased with depth, with concentrations generally <1 mM throughout Unit II. Phosphate ranged from 13.5 to 3.6 µM in Unit I and is mostly below detection in Unit II. The elevated ammonium concentrations were unusual and were even greater than those observed in the interstitial water samples of the WRCs.

Calcium and magnesium concentrations decreased with depth in a fashion similar to the drilling mud. Magnesium concentrations (33.5–10.6 mM) are greater than those of the drilling mud, with the exception of one sample at the top of Unit II (0.15 mM). Calcium concentrations (2.44–0.76 mM) were lower than the drilling mud associated with Units I and II, with the exception at the top of Unit II (62.06 mM). As with the mud water samples, cuttings water samples presented significant amounts of dissolved barium (4.3–12.6 µM).

The differences between squeezing the deepest cuttings water samples (337-C0020A-187-SMW, 197-SMW, and 213-SMW) and those collected using a Rhizon sampler (197-SMW) were minimal. In this case, the Rhizon sampler was used with wet cuttings and no attempt was made to remove drilling mud from the cuttings before sampling, such that the fluid that was collected was in contact with the outside of the cuttings. The results with the Rhizon sampler for salinity, alkalinity, sodium, and potassium were bracketed by the values from the two adjacent squeezed cuttings, whereas results for chloride, bromide, and sulfate are similar to the values of one of the adjacent squeezed samples. These results suggest that the predominant source of water extracted by squeezing of the cuttings is actually the fluid on the exterior surfaces of the cuttings and not the fluid within. Dissolved ions of the alkaline earth elements Mg2+ (15.1 mM), Ca2+ (4.3 mM), Ba2+ (11.7 µM), and Sr2+ (54.2 µM) were all significantly greater than the adjacent squeezed counterparts. Although the cause is not clear, it may be that the samples taken with the Rhizon sampler were less prone to precipitation loss of carbonates and sulfates in the brief time that it took to extract the water.

Overall, the cuttings data provide limited information regarding the in situ interstitial water composition of Site C0020 because of the high degree of contamination from drilling fluid. They do, however, provide an end-member for any interactions that might occur between the drilling mud and extremely contaminated sediment. Enrichments in ammonium observed in the cuttings water samples of Unit I were greater than those observed in the interstitial water samples. The enrichment of ammonium is likely due to desorption of exchangeable ammonium caused by contact with the drilling mud, which is highly enriched in KCl (see above); a common analytical method to determine exchangeable ammonium in soils is to leach them with 1–2 M KCl solution such that K+ saturates all of the NH4+ adsorption sites (Kowalenko and Yu, 1996).

Whole-round cores

A total of 48 whole-round samples were collected from 32 cores for interstitial water analyses in Hole C0020A, and the whole-round lengths ranged from 15 to 76 cm. Because of low porosity and permeability (see “Physical properties”), only 24 of the processed WRC samples yielded interstitial fluid through squeezing, with low volumes between 33.5 and 0.2 mL despite large subsamples (up to 70 cm per WRC for selected segments). Of the 24 cores that yielded interstitial water, 15 were from sediment in Unit II, 7 from Unit III, and 2 from Unit IV.

The interstitial water yield is shown in Figure F25, expressed as a volume of the fluid extracted per volume of sediment squeezed. The sediment volume was calculated based on the length of the whole-round section and the diameter of the section after peeling. The greatest yields (>30 mL/dm3) were in poorly consolidated sandstone, including highly contaminated samples from the top of Unit II. Silty sandstone located in Units II and III had variable yields (0–32 mL/dm3). In Units II and III, yields from siltstone were 0–12 mL/dm3. No interstitial water was gained from coal and shale samples regardless of the location. In Unit IV, very little fluid (0.2–1 mL/dm3) if any was extracted, regardless of lithology.

When cleaning off the cores for interstitial water squeezing, it was observed that after scraping off the potentially contaminated drilling mud fluid many of the clay-rich samples initially appeared dry underneath and then became wet again after a few seconds. This might have been due to capillary action or was possibly caused by gas separation from the fluids that pushed the interstitial water from the center toward the outside of the cores.

Salinity, total alkalinity, and pH

Downhole salinity information was derived based on the refractive index data collected for the 24 analyzed whole-round samples (Fig. F26A; Table T14). A large range in salinity values from 27‰ to 77‰ is observed throughout the sequence in Hole C0020A. The most saline fluids (68‰–77‰) were situated at the top of Unit II, where pore water was highly contaminated with saline drilling mud and should not be taken into consideration when describing downhole trends in pore water chemistry. Neglecting these highly contaminated samples, the salinity remained at ~40‰ in Units II and III and decreased at the bottom of Unit III, reaching 27‰ in Unit IV. Salinity values of ~40‰ were probably still significantly impacted by contamination. Nevertheless, the decreasing trend with depth is consistent with a gradual change in the paleosedimentary environment and may reflect the influence of a low-salinity end-member, suggesting an estuarine environment (see “Lithostratigraphy” and “Paleontology”). Alternatively, the presence of a deeper pure water source (e.g., clay dewatering, etc.) may have resulted in fluid migration, characterized by a lower salinity, in the overlying sediment column.

Interstitial water samples from 335 mbsf analyzed during the Chikyu shakedown cruise (Expedition CK06-06) revealed a maximum alkalinity of 127 mM at 299.6 mbsf, followed by a decrease in alkalinity to 77 mM at 360.8 mbsf (Tomaru et al., 2009; Higuchi et al., 2009). Excluding the highly contaminated samples from the upper part of Unit II (Fig. F26B; Table T14), the highest alkalinity was 14.2 mM found in Unit III. This indicates that pore water alkalinity decreases significantly with depth and most likely indicates that active contributions of dissolved inorganic carbon to the alkalinity through processes such as anaerobic oxidation of methane (AOM) (e.g., Borowski et al., 2000) are much less significant in Units II–IV than in the shallower sediments investigated in the shakedown cruise.

Given that the yields of interstitial water in the deeper analyzed parts of Hole C0020A were limited and leaving heavily contaminated samples out of consideration, only a few valid pmH measurements were carried out. At the top of Unit II, pmH values of 8.3–8.6 were found, followed by a pmH of 9.1 at the bottom of this unit. In Unit III, pmH values range between 7.5 and 8.2 (Fig. F26C; Table T14).

Chloride, bromide, and sulfate

Excluding the contaminated samples at the top of Unit II, chloride concentrations of interstitial water in Hole C0020A showed relatively little change with depth, with average values ~600 mM (Fig. F27A; Table T14). Positive deviations (as high as 715 mM) were likely a result of drilling mud contamination. In general, chloride concentrations were higher than those observed during the Chikyu shakedown cruise, which decreased very gradually in a linear fashion from 542 mM at 1.4 mbsf to 525 mM at 363.1 mbsf (Tomaru et al., 2009; Higuchi et al., 2009). The lowest values for Hole C0020A were 488 and 459 mM from Units III and IV, respectively. Whether these two values represent local freshening associated with the coalbeds or are part of the gradual decreasing trend, presumably due to clay alteration at depth (e.g., Tomaru et al., 2009), is uncertain.

Bromide exhibits a similar pattern, with concentrations fluctuating at ~800 µM throughout the analyzed sequence (Fig. F27B; Table T14). This value roughly coincides with the shallow pore fluid from the Chikyu shakedown cruise, which yielded a Br concentration of 824 µM at 1.4 mbsf, yet is significantly lower than Br concentrations that are in the deeper subsurface sediment (150–363 mbsf), which are generally 1000–1200 µM (Tomaru et al., 2009). As for the case of chloride, a decrease in bromide concentrations is observed at the bottom of Unit IV at 459 µM.

The dissolved sulfate concentrations across the analyzed sequence illustrate the extent of drilling mud water incorporation into the sediment cores (Fig. F27C; Table T14). In the case of highly contaminated samples, sulfate concentrations in the interstitial water samples were as high as 9.9 to 20.2 mM at the top of Unit II. Below these uppermost samples in Unit II, sulfate concentrations at range from 0.97 to 2.31 mM. This is somewhat greater than the majority of samples collected during the Chikyu shakedown cruise in 2006 from below the sulfate–methane transition, which exhibited sulfate concentrations either below detection or between 0.1 and 0.3 mM, with the exception of sampled sandy intervals that showed contamination by modern seawater and sulfate as high as 11 mM (Tomaru et al., 2009). The increase in sulfate in Units III and IV to values of 1.1 to l1.35 mM corresponds to an increase in sulfate in the drilling mud.

Phosphate and ammonium

The recorded phosphate and ammonium concentrations of the interstitial water samples can be assumed to represent values for in situ interstitial water concentrations diluted somewhat by drilling mud because there are only two exceptions where drilling mud has detectable amounts of these two species. Phosphate was not detectable throughout the major part of the sequence, with the exception of two interstitial water samples in the uppermost part of Unit II (Table T14). These samples were characterized by phosphate concentrations of 3.3–5.4 µM, which are close to the detection limit of ~2.7 µM.

Ammonium was detectable throughout Hole C0020A (Fig. F28A; Table T14). Ammonium concentrations were 3.3–7.8 mM in the upper, most contaminated samples of Unit II and decreased somewhat to between 2.0 and 2.89 mM in the lower part of the unit and 1.3–3.2 mM in Unit III. The data possibly suggest a downhole decrease in interstitial ammonium concentrations; however, there is no certainty regarding the actual source. The measured concentrations might represent (1) a diluted signal after incorporation of drilling mud water into the interstitial water sample, (2) natural variation in the interstitial water composition, or (3) variable desorption from clay minerals and organic matter exposed to the highly alkaline drilling mud.

Interstitial water samples were also analyzed for nitrate using ion chromatography. Reoccurring values of around a few hundred micromolar (maximum = 440 µM) were found in roughly half of the pore water samples; however, all of the values were below or close to the method’s detection limit (350 µM). Accordingly, the nitrate results are considered to be unreliable and, therefore, these data were not included in the results table.

Alkali metals and alkaline earth elements

Sodium concentrations generally paralleled those of chloride, with relatively little change throughout the column (Fig. F28B; Table T14). The majority of the samples fell in the range of 515–620 mM. Several high values in Units III and IV (630 to 667 mM) correspond to high chloride values and probably are influenced by the addition of NaCl to the drilling mud over this interval. Low values of 466 and 413 mM in Units III and IV correspond to the lowest chloride values, which, as discussed, may represent either freshening through clay dewatering or the influence of an estuarine depositional environment.

Potassium was one of the major components of the drilling mud because of the addition of KCl and KOH. The admixture of even minor amounts of mud water with interstitial water has a serious impact on the overall K+ concentrations (Fig. F28C; Table T14). For example, a 5% mixture of mud water (Ca2+ = 5.95 mM, K+ = 847 mM) with seawater (Ca2+ = 10.55 mM, K+ = 10.44 mM) yields a net composition of 10.3 mM Ca2+ and 52.3 mM K+. High levels of contamination were clearly found at the top of Unit II, where potassium was as high as 448 mM. In cases where the contamination was not as high, values generally ranged from 10 to 50 mM, and in several instances within Units III and IV, potassium was actually below seawater values (2.34–3.31 mM).

Magnesium concentrations were generally greater in the interstitial water samples than in drilling mud at similar depths (Fig. F29A; Table T14). The uppermost four samples of Unit II show levels of magnesium (31.8–47.0 mM) that exceed those of the drilling mud water in the same unit (14.0–23.8 mM). The least contaminated samples in Units III and IV have concentrations of 12.4 and 7.3 mM, respectively, compared to 2.2–5.2 mM in the drilling mud across the same intervals. Although the interstitial water magnesium values are significantly greater than the mud water values, they are lower than concentrations measured during the Chikyu shakedown cruise for deep sediment interstitial water, which all exceed 70 mM at depths >200 mbsf (Higuchi et al., 2009). The loss of magnesium in the interstitial water is likely due to authigenic mineralization of dolomites between sand grains, which was observed in some of the deeper sediment layers associated with the coalbeds (see “Lithostratigraphy”).

Calcium concentrations in the interstitial water were high (14.7–39.2 mM) throughout the sediment column (Fig. F29B; Table T14) and were significantly greater than drilling mud concentrations (3.5–8.4 mM). They also exceed the interstitial water calcium concentrations of ~4 mM observed in most samples collected deeper than 300 mbsf during the Chikyu shakedown cruise (Higuchi et al., 2009). High calcium concentrations at Site C0020 may indicate that the low-alkalinity water in Units II–IV is undersaturated in calcite or aragonite but still saturated with regard to dolomite. This would suggest that progressive dolomitization is occurring in calcite as it is buried. Dolomites are often associated with microbial activity and AOM in organic-rich sediments along continental margins (e.g., Meister et al., 2007).

Strontium concentrations in the interstitial water (93.9–719.6 µM) also greatly exceed concentrations found in drilling mud (13.6–57.8 µM), and thus the influence by contamination on this element is probably low (Fig. F29C; Table T14). Strontium concentrations uniformly exceed seawater concentrations (~93 µM) as well as in the deeper cores from the Chikyu shakedown cruise (~90 µM at 200 mbsf to ~80 µM at 300 mbsf) (Higuchi et al., 2009). Because strontium is generally coprecipitated with calcium in authigenic calcites to a much greater degree than in authigenic dolomites (Snyder et al., 2007), the increase in strontium concentration may be another indicator of deep dolomitization.

Barium concentrations are low at the top of Unit II (12.8–31.1 µM), are greatest in Unit III (29.25–78.76 µM), drop in Unit IV (25.1–38.9 µM) (Fig. F29D; Table T14), and are, in general, greater than drilling mud values (5.7–14.43 µM, with the exception of one mud sample at 42.6 µM in Unit II). The elevated values in the drilling mud probably represent in situ dissolution of sedimentary barite in the pore water under anoxic conditions (e.g., Torres et al., 1996). Because of contamination by drilling mud and resulting sulfate enrichment, the sampled interstitial water is extremely supersaturated with respect to barite. However, precipitation is either inhibited by other components in the mud or the reaction kinetics are such that there is not enough time for a substantial portion of the dissolved barium to precipitate. In Unit IV, barium concentrations are roughly half those of Unit III, but given the presence of drilling mud sulfate, it is not possible to determine whether this is caused by a drop in the in situ pore water Ba concentrations or if more barium simply precipitated from the pore water during sediment recovery and squeezing.

Transition metals and metalloids

Boron concentrations are generally somewhat lower than seawater (416 µM) in Units II–IV (Fig. F30A; Table T14). One sample was as low as 82 µM and four other samples were <150 µM, but generally the interstitial water samples presented concentrations comparable to or greater than the concentrations of boron in mud water samples (166–238 µM). Two samples in Unit II were greater than seawater at 465 and 691 µM, whereas one sample in Unit IV was significantly enriched at 1313 µM. The enrichments may be indicative of the diagenetic release of boron from sedimentary organic matter (e.g., Snyder et al., 2005; You et al., 1993). Desorption and adsorption on clay minerals in the alkaline drilling mud likely has an influence on the squeezed pore water, such that the variations in observed boron concentrations may not be representative of the actual in situ composition of interstitial pore water.

Lithium enrichments may be indicative of the dewatering reactions in deeply buried clay (Martin et al., 1991). Several cores from Unit III presented lithium concentrations >300 µM, which is higher than the drilling mud across the lower part of Unit II through Unit III (151.8–259.8 µM); however, the most contaminated interstitial water sample in the upper part of Unit II also contained lithium at 353.3 µM (Fig. F30B; Table T14). Although all of the samples showed significant enrichments relative to average seawater (26 µM), the overlap with drilling mud concentrations makes it difficult to derive more than a general observation that pore water does show some degree of enrichment in Unit III.

Dissolved silica concentrations generally decrease with depth from as high as 867 µM in Unit II to as low as 68.9 µM at the deepest sample recovered from Unit IV (Fig. F30C; Table T14). In contrast, dissolved silica concentrations in the mud samples increase with depth either because of the dissolution of silica-bearing fossils (see “Paleontology”) or the dissolution of silica in the ASTEX-S SAS additive (Eikenberg, 1990; Southwick, 1985). The lower silica concentrations in deeper interstitial water samples may be due to lower contamination from drilling mud.

Interstitial water tends to have less iron than the drilling mud (Fig. F30D; Table T14), and thus, the water samples are highly sensitive to iron contamination. The highest iron concentrations (5.1–25.0 µM) are in the top of Unit II, which makes the actual interstitial water composition suspect. In contrast, manganese concentrations in interstitial water (Fig. F30E; Table T14) exceed those in the drilling mud, particularly in the coal-bearing Unit III, where the maximum dissolved manganese concentration is 41.3 µM. However, the long squeezing times of some samples under oxic conditions may have oxidized both iron and manganese to their higher, less soluble oxidation states (Stumm and Lee, 1961) so that actual concentrations may have been higher within the formation than the observed values of squeezed pore water. The increase in Mn across Unit III roughly parallels the enrichments observed in Li.

Formation water samples

This is the first IODP expedition in which a downhole Quicksilver In Situ Fluid Analyzer (IFA) was deployed to select and retrieve fluids directly from the walls of the borehole, collecting them in a single-phase multisample chamber (SPMC) and then transferring them shipboard into Single-Phase Sample Bottles (SSB) (see “Wireline logging” in the “Methods” chapter [Expedition 337 Scientists, 2013b]). In preparation for IFA sampling, a decision was made to seek out the sandy layers that were in close association with the coalbeds for sampling. The rationale behind this decision was that the high permeability of the sand would allow the fluid flowline of the IFA to more effectively draw fluids from deeper within the target sediment in a reasonable amount of time. During sampling, the guard flowline would allow any surficial fluids that are contaminated by drilling mud to be effectively drawn away from the point of contact between the sediment formation and the fluid flowline (Mullins, 2008).

Deployment of the IFA was successful in precisely identifying potential sampling sites down the borehole, and Downhole Fluid Analysis (DFA) provided a preliminary assessment of the permeability and flow of formation fluid into the probe. Six samples were collected in the SPMC from Units II and III (see “Wireline logging” in the “Methods” chapter [Expedition 337 Scientists, 2013b]) and transferred to SSBs for gas and water analysis. Unfortunately, the formation water samples all showed drilling mud contamination, including high salinity and high chloride (Table T15). Sulfate concentrations in the formation water ranged from 7.9 to 23.9 mM, which was less than drilling mud water across Units II and III (20.52–55.96 mM) but only slightly lower than highly contaminated drilling cuttings across these intervals (18.6–23.3 mM) and greater than the pore water of lower part of Unit II and all of Unit III (0.97–10.5 mM). Potassium in the SSBs also showed significant drilling mud contamination, with concentrations ranging from 76.5 to 375.5 mM, which is less than mud water samples across these units (643.0–732.9 mM) and less than water extracted from drilling cuttings in Unit II (648.4–781.5 mM) but is comparable to the highly contaminated pore water samples at the top of Unit II (65.2–448.5 mM) and is significantly more contaminated than the samples from the lower part of Unit II and Unit III (3.3–57.1 mM).

Despite these results, a number of interesting conclusions can be made that will be of some use in the future design and deployment of in situ sampling devices. In some respects, formation water shows a change in composition between Units II and III. Sodium increases from 550–567 mM in Unit II to 641–861 mM in Unit III (Fig. F31A). Sulfate concentrations are between 5.1 and 7.9 mM in Unit II and more enriched in Unit III, with values from 11.4 to 23.8 mM (Fig. F31B). Chloride increases from 670–777 mM in Unit II to 769–1125 mM in Unit III (Fig. F31C). This change is likely due to the change in the formulation of the drilling mud between Units II and III to include ASTEX-S SAS as well an increase in both sodium chloride and potassium chloride. Despite the high sulfate concentrations, dissolved barium is present in the formation water (12.1–52.7 µM) in concentrations similar to the pore water over the lower parts of Unit II and Unit III (9.0–78.8 µM) and greater than the mud cuttings in Unit II (4.6–11.7 µM), suggesting that either the drilling mud contamination occurred at the time of sampling and the samples did not have time to precipitate barite, or barite precipitation was inhibited by other ions in solution.

Contamination assessment

The degree to which the sediment pore water has been contaminated by drilling mud may be determined quantitatively if a number of working assumptions are made. An estimate of the fraction of mud water contamination in the pore water may be determined by assuming binary mixing between mud fluid and interstitial water. In order to determine this fraction, a single ionic species whose composition is distinct between the mud end-member and a pore water end-member must be selected, where the change in composition in the uncontaminated pore water is assumed to be negligible with depth. Finally, it is assumed that the mixing process is essentially nonreactive. The degree to which these assumptions are considered valid or not is evaluated by using several different ionic species and comparing the calculated fraction of mud water contamination. In this particular case,

XDML = ([A]meas – [A]FW)/([A]DML – [A]FW),


  • XDML = fraction of drilling mud liquid mixed into the sampled water,

  • [A]meas = measured concentration of species A,

  • [A]FW = assumed concentration of species A in the formation water, and

  • [A]DML = concentration of species A in the drilling mud.

In the case of interstitial water, drilling cuttings, and formation water, XDML was determined using three different parameters: sulfate, potassium, and salinity because within the Chikyu shakedown cruise Hole C9001C (Tomaru et al., 2009; Higuchi et al., 2009) the formation water concentrations ([A]FW) of these three parameters did not appear to change appreciably below 200 mbsf. The assumption is made that [SO42–]FW is 0 mM because samples were collected well below the sulfate–methane transition (4.5 mbsf). Based on the deepest sample (363 mbsf) from this previous cruise, [K+]FW is 12.2 mM and [salinity]FW is 38‰. Measured interstitial water concentrations ([SO42–]meas, [K+]meas, and [salinity]meas) from Table T14 were matched with the mud composition from the nearest depth within the same unit in order to select the appropriate values of [SO42–]DML, [K+]DML, and [salinity]DML from Table T12. Using this approach, the mud fraction was calculated and plotted with depth (Fig. F32). The average and standard deviation between the three derived mud fractions are also shown in the plot for the formation water.

The mud fraction in the interstitial water, based on sulfate in the most contaminated upper part of Unit II, ranged from 0.43 to 0.88 and then generally decreased to below 0.1. The mud fraction based on potassium was 0.32–0.69 in the contaminated samples in the upper part of Unit II and was generally <0.05 throughout the rest of Unit II and all of Units III and IV. Based on salinity, the contaminated upper samples had mud fractions between 0.50 and 0.65, whereas the deeper samples generally had mud fractions of 0.05 or less. When calculating the mud fraction from potassium and salinity, a few values were <0, suggesting actual shifts in the formation water potassium and salinity that make the estimated [A]FW value larger than it should be.

Using this approach for the cuttings water, showed even higher degrees of contamination based on sulfate concentrations, with mud fractions ranging from 0.76 to 0.90 in Unit II and 0.81 to 1.00 in Unit III. The mud fraction based on potassium could not be used because potassium concentrations in the cuttings were greater than both those in the pore water and those in the drilling mud, presumably caused by clay dissolution by the alkaline mud. The mud fraction in the cuttings based on salinity was 0.690–0.75 in Unit II and 0.81–1.00 in Unit III.

The mud fraction in the formation water collected in the SSBs was 0.25–0.35 in Unit II and 0.20–0.43 in Unit III, based on the presence of sulfate. Based on potassium concentrations, the fraction of mud was 0.10–0.32 in Unit II and 0.17–0.51 in Unit III. Finally, the salinity of the formation water suggests a range of mud fractions of 0.03–0.24 in Unit II and 0.09–0.33 in Unit III. (Fig. F32). With the exception of the shallowest SSB in Unit I, all other samples show greater fractions of mud water contamination than are observed in the squeezed pore water samples. Using the mud fractions derived from sulfate, potassium, and salinity and the concentrations of the other dissolved species for each sample in Table T14, corrections in the pore water concentrations could be carried out using

[A]corr = ([A]measXDML[A]DML)/XIW,


  • [A]corr = corrected in situ concentration of the analyte,

  • [A]meas = measured concentration of the analyte,

  • [A]DML = element concentration in the drilling mud water,

  • XDML = molar fraction of drilling mud liquid introduced into the sample during drilling, and

  • XIW = fraction of in situ interstitial water in the sample, and where XDML + XIW = 1.

No attempt was made to carry out corrections where XDML > 0.2.

The approach was taken of averaging the three corrections (sulfate, potassium, and salinity) and taking the error between the three to be 1 standard deviation from the mean. Results are shown in Figure F33 for alkalinity, Ca, Mg, and Sr. Note that error bars for alkalinity are much larger than for Ca, Mg, and Sr because the mud has an alkalinity similar to the interstitial water samples. In contrast, errors for the alkaline earth elements are relatively small because the mud concentrations for these elements are significantly lower than those of the interstitial water. The increase in alkalinity, Ca, Mg, and Sr in the corrected profiles of Figure F33 becomes much more discernible in the corrected profiles than in Figures F26 and F29.