蔡春芳

 

 

论文题目：流体地球化学:成因、流动与流体-岩石相互作用及塔里木盆地实例分析   

 

作者简介：蔡春芳，男，1966年03月出生，1997年09月师从于中科院地质与地球物理研究所汪集旸教授，于2000年08月获博士学位。

                                       

摘  要

 

盆地地质流体与油气生成、运移、聚集和保存紧密相关，属于地学前缘研究课题。利用多学科交叉方法，进行了长石在有机酸、原油、高矿化度卤水中的溶解实验，研究了长石溶解后元素Al在油水相中的分配；通过详细研究塔里木盆地塔中地区、和田河气田、库车坳陷和塔北隆起北部油气藏流体的化学、同位素组成，观察岩石薄片、测试流体包裹体成分和均一化温度、全岩和自生矿物同位素组成，并参考和结合油气运移和成藏史的研究成果，分析了三地区流体的成因、流动和有机-无机相互作用，包括烃类侵位对成岩作用的影响、微生物和热化学的硫酸盐还原作用。并对比研究了微生物和热化学硫酸盐还原作用发生的条件和产物。

以前的研究显示，油田水中Al浓度（多低于3mg/L）远低于储层矿物溶解所预期的值，同时，世界范围内油田水中有机酸实测结果均以乙酸和乙酸根为最主要的羧酸和其阴离子，约占总有机酸的70%以上。与草酸不同，乙酸对矿物的溶解性以及络合Al和Si能力尚存争议。为探讨该问题，实验配制了含2mol/L NaCl、0.05mol/L乙酸、5ml原油、0.05mol/L草酸(这些值接近塔里木盆地油田水实测值)，并控制初始pH₀、温度，进行了16种不同条件下长石溶解作用的模拟实验研究。实验结果表明，在含原油的溶液中，部分Al分配到油相中，且在含羧酸的溶液中，进入油相中Al的比例增大。在含草酸的水溶液中，Al和Si的浓度明显高于含乙酸的溶液，与稳定的草酸硅和草酸铝配位化合物形成有关。乙酸根可增高Ca、Al的活性，但Al-乙酸根间的配位化合物所占的比例较低。

油田水的成因复杂，本论文测试了水化学和同位素组成来揭示其成因和水—岩反应的信息。对于同沉积水为海水的储层而言，利用油田水Cl与Na、Ca,、Mg、 Sr、Br、K间相对于海水蒸发曲线的关系，结合d¹⁸O — dD关系研究了油田水的成因。认为塔里木盆地塔中古生界、和田河气田油田水是海水蒸发残余，后期受到了淡水混合、岩盐溶解、方解石白云石化、长石的钠长石化等水-岩反应的影响。d¹⁸O和dD测试数据则显示，淡水是沿和田河气田石炭系和奥陶系不整合面由西向东流动的。这与前人的观点正相反。而对于沉积在淡水环境下的储层而言，主要通过摩尔Na/Cl比值、Cl离子浓度、d¹⁸O — dD关系来研究其成因与演化。d¹⁸O — dD关系揭示了，库车前陆盆地第三系和白垩系油田水为淡水蒸发后又与现代淡水混合的产物，后期经岩盐溶解改造成因，淡水能通过断层等通道下渗到寒武—奥陶系。

⁸⁷Sr/⁸⁶Sr—1/Sr关系可以很好地应用于研究碳酸盐岩层（低⁸⁷Sr/⁸⁶Sr、高Sr）、碎屑岩层（高⁸⁷Sr/⁸⁶Sr、低Sr）和基岩（高⁸⁷Sr/⁸⁶Sr、高d¹⁸O）中卤水间的流动与混合作用。应用这种方法，查明塔中地区寒武—奥陶系碳酸盐岩储层富Sr、贫⁸⁷Sr的油田水向上流动，并与志留系和石炭系砂岩储层贫Sr、富⁸⁷Sr的油田水发生了混合。这与目前油气勘探所揭示的油气运移方向是一致的。由北西向南东方向，奥陶系、志留系和石炭系油田水⁸⁷Sr/⁸⁶Sr比值均降低，可能指示了北西部阿瓦提泥页岩中富⁸⁷Sr的流体，向南东方向流动和混合。志留系油田水明显受到淡水的混合，使Sr浓度降低，但对⁸⁷Sr/⁸⁶Sr比值影响较小。和田河气田与之类似，奥陶系碳酸盐岩中的卤水（低⁸⁷Sr/⁸⁶Sr、高Sr）向上运移，并与石炭系油田水发生混合作用。库车前陆盆地寒武—奥陶系碳酸盐岩中油田水⁸⁷Sr/⁸⁶Sr高于碎屑岩储层，结合气体d¹⁵N 、³He/⁴He数据，认为来自基岩。 

烃类侵位对成岩作用的影响尚存分歧，但具有重要的研究价值。如果知道某些成岩作用因油气的侵位而停滞，便可望据此确定油气成藏的相对或绝对年龄。同时，油气侵位是否有利于次生孔隙的发育，也值得研究。塔中地区志留系实例研究显示，烃类侵位后可因淡水注入而被氧化和生物降解，所产生的酸性环境有利于矿物的溶解。烃类侵位还使成岩介质还原性增强，胶结物富含还原的2价铁；但未能阻滞石英的次生加大，可能与含油饱和度低有关。水湿石英表面硅的沉淀而产生的浓度梯度，导致发生扩散作用，提供了石英次生加大作用所需Si的来源。

有机质的氧化-硫酸盐还原作用是油气藏和源岩中一种重要的有机-无机相互作用类型，不仅可使油气或源岩中有机质发生次生蚀变作用，而且所产生的H₂S、有机CO₂可以改善储层物性。本论文研究了两案例，其中H₂S分别为热化学硫酸盐还原作用和微生物硫酸盐还原作用成因。我们发现，塔里木盆地塔中地区古生界储层中相对富含H₂S 和CO₂。为了评估H₂S的成因，分析了气体、矿物胶结物和全岩的组分和同位素，并进行了岩石学和流体包裹体的研究。由于N₂和He没有幔源的同位素特征，H₂S 和CO₂不可能来自深部地壳。同时，根据硬石膏被交代的起始深度、包裹体均一化温度和埋藏—热史数据，认为方解石交代硫酸盐矿物发生在>125ºC的地温条件下。而储层温度>125ºC，显然太高，不适合于微生物的生长，所以也不可能来自微生物的硫酸盐还原作用。寒武系—下奥陶统碳酸盐岩烃类与硬石膏间的热化学硫酸盐还原作用是最可能的来源。自生黄铁矿的d³⁴S接近于同时代地层中硬石膏的同位素值，CO₂和方解石胶结物的d¹³C轻达-21‰PDB，支持了H₂S和CO₂的为热化学硫酸盐还原成因。H₂S 和CO₂通过穿层流动进入了志留系储层中。其中部分H₂S已经不再以H₂S分子存在，而广泛地生长成岩晚期黄铁矿。同样地，部分CO₂因成岩晚期方解石的沉淀而不再以流体相存在。

而微生物硫酸盐还原作用是自然界中更普遍的硫酸盐还原作用。一般认为，硫酸盐还原菌（SRB）以有机酸、醇等为碳源和能源来生长和繁殖的，而不能直接利用烃类。就本人所知，仅在1999年在深海钻探采集的样品中，发现甲烷能被SRB直接消耗。本论文对塔里木盆地和田河气田研究显示，SRB能同时利用有机酸和烷烃气，并选择性地利用¹²C 的烷烃气，导致残余的烷烃气的d¹³C变重。具体地，在和田河气田石炭系和奥陶系不整合面（C/O）附近气藏中检测出高的H₂S（可达2000ppm）和CO₂浓度。该H₂S和CO₂不可能来自幔源，因为³He/⁴He并没有幔源的特征。从东部向西部，H₂S浓度和CO₂/( CO₂ +ΣC_1-4) 增高，烷烃气的d¹³C 变重，其中甲烷d¹³C由-38.6‰变为-35.6‰。然而，H₂S不可能来自热化学硫酸盐还原作用。因为该气藏目前储层的温度太低，仅为60-80ºC，在深部储层中本应该出现更强的热化学硫酸盐还原，却没有检测出H₂S，黄铁矿也少见，而且，成岩晚期自生黄铁矿的d³⁴S 轻达-24.9‰，于是推测H₂S的d³⁴S值远轻于奥陶系硬石膏（约+26‰）和已知有机质最轻的d³⁴S值（-17‰），表明研究区H₂S不可能来自TSR，也不可能来自有机质的分解。而最可能来自微生物硫酸盐还原作用。SRB除了消耗有机酸外，还直接或间接地利用了烷烃气。随着硫酸盐还原程度（CO₂/( CO₂ +ΣC_1-4)）的增强，残余烷烃气d¹³C 变重。同时，相对于乙烷、丙烷而言，SRB选择性地消耗甲烷，于是，甲烷d¹³C 正偏移的程度最大。在古老地层中，不可能有沉积时残留的SRB。于是，SRB的来源至关重要。地层水氢、氧稳定同位素关系和水化学分析揭示了，在C/O不整合面附近，具有现代淡水的混合作用。硫酸盐还原菌是由在西部入口的淡水携带进入储层的，并沿不整合面由西向东流动。因而，西部气藏中硫酸盐还原菌数量最多，还原作用最强。结果，西部CO₂的d¹³C 最轻（轻达-24.3‰），烷烃气的d¹³C 变重最明显。于是甲烷d¹³C —R_o（源岩镜质体反射率）的经验方程在研究区不适用了。
　

 Abstract

   Geofluid in sedimentary basins is related to petroleum generation, migration, accumulation and preservation, and is a topic of geological frontier. By integrating the multi-discipline methods, the thesis has carried out experiments to dissolve microcline in solutions with organic acids, crude oil, brines with high total dissolved solids (TDS) to determine Al distribution between the crude oil and brines. Cases for the study include Central Tarim, Hetianhe Gas Field and Kuche forland basin with data containing fluid chemistry and isotopic compositions, thin sections of sandstones and carbonates, homogenization temperatures and salinities of fluid inclusions, isotopic composition of bulk rock and authigenic minerals. The aims are to elucidate fluid origin and flow and effect of hydrocarbon emplacement on diagenesis, and to show occurrence of bacterial and thermochemical sulfate reduction in the three areas of the Tarim Basin.

Previous studies show that Al concentrations in oilfield water are much lower than expected from the dissolution of minerals observed in thin sections. Analyses of organic acids and anions (OAA) in oilfield water show that acetic acid and acetate are the most important type of OAA, and generally constitute >70% of OAA in the Tarim basin and other basins. Different from oxalic acid, there exist some debates on dissolution of feldspar under solutions with acetic acid and on how important for acetate complexing of Al and Si. To explore the issue, 16 experiments to dissolve microcline have been conducted using different solutions with 2mol/L NaCl, 0.05mol/L acetic acid, 5ml crude oil or 0.05mol/L oxalic acid (the values are close to those measured in Tarim oilfield waters) and different initial pHs and temperatures.  One of the results shows that after 100 hours, part of the dissolved Al was detected in the crude oil, and the Al concentrations in the crude oil rose when acetic or oxalic acids are added. The result can be used to explain that most oilfield waters in the Tarim Basin are characterized by less than 3mg/L Al. Crude oil added to the solutions can enhance microcline dissolution, which is also observed in the case – Silurian sandstones with early emplacement of crude oil in the Central Tarim. Al and Si have higher concentrations in the solution with oxalic acid than in solution with acetic acid, suggesting that Si-oxalate and Al-oxalate complexes are more significant in amount than Si – acetate and Al-acetate complexes. Presence of acetate can enhance the activity of Ca and Al, but Al concentrations have not been increased significantly due to limited Al-acetate complex formed during the experiments.

The origin and chemical evolution of oilfield waters may involve a variety of processes. Chemistry and isotopic composition of oilfield waters from the Tarim Basin have been measured to elucidate the origin and water-rock interaction. As for oilfield water from the reservoirs deposited in marine environment, relationships between Cl and Na, Ca, Mg, Sr, Br and K relative to seawater evaporation trajectory in conjunction with relationship between dD and d¹⁸O can be used to account for the origin of oilfield waters. The results show that the waters from Paleozoic reservoirs in Central Tarim and from reservoirs near the Carboniferous/Ordovician unconformity are evaporated seawater waters, which subsequently mixed with meteoric water, and were influenced by water-rock interactions such as salt dissolution, dolomitization of calcite, albitization of feldspar. Contrary to the previous study, dD and d¹⁸O measurement in the study shows that meteoric water flowed eastward along the Carboniferous/Ordovician unconformity in the Hetianhe Gas Field. As for oilfield waters originated from meteoric water, molar Na/Cl ratios, Cl concentrations, and dD and d¹⁸O are used to explain the evolution. The relationships between dD and d¹⁸O and between molar Na/Cl ratio and Cl concentration show that the oilfield waters from the Tertiary and Cretaceous in Kuche forland basin are evaporated meteoric water, which subsequently mixed with modern meteoric water, and that salt dissolution significantly alter water chemistry. The modern meteoric water may have permeated Cambrian – Ordovician strata along conduits such as faults.

Relationship between ⁸⁷Sr/⁸⁶Sr and 1/Sr can be used to indicate migration and mixing of brines from carbonate strata (low ⁸⁷Sr/⁸⁶Sr ratio but high Sr content), clastic strata (high ⁸⁷Sr/⁸⁶Sr ratio but low Sr content) and crystalline basement (high ⁸⁷Sr/⁸⁶Sr ratio and heavy d¹⁸O value). Using this approach, it can be found that ⁸⁷Sr-depleted brine from Ordovician carbonates may have migrated up to and mixed with ⁸⁷Sr-enriched waters from Silurian and Carboniferous sandstones, and that Silurian brines have mixed with meteoric water. The result is consistent with that of the present-day petroleum exploration. From northwest to southeast, oilfield waters from the Ordovician, Silurian and Carboniferous show decreased trends of ⁸⁷Sr/⁸⁶Sr ratio, indicating that ⁸⁷Sr-enriched water from shale and mudstones in the Awati to the northwest might flow toward southeast and mixed with previously low ⁸⁷Sr water. Oilfield water in the Silurian reservoirs may have mixed with meteoric water, which significantly decreased Sr concentrations but no significant change in ⁸⁷Sr/⁸⁶Sr ratio. In the Kuche forland basin, brines from the Cambrian and Ordovician carbonates have higher ⁸⁷Sr/⁸⁶Sr ratios than those from the overlying sandstones, when combined with chemistry, d¹⁵N and ³He/⁴He ratios of the coexisting natural gases, suggesting that the brines were derived from the basement.

There exists some debate on the effect of hydrocarbon emplacement on mineral diagenesis. Case-study from Silurian sandstones in the Central Tarim show that quartz has kept overgrowing secondarily when oil saturation was decreased by meteoric water flushing subsequently to hydrocarbon emplacement. Silicon precipitates on the water-wet quartz surface, leading to decreased Si concentration close to the surface. A silicon concentration gradient can result in silicon diffusion, which supplies Si for quartz overgrowth.

   Hydrocarbon oxidation-sulfate reduction is an important type of organic-inorganic interaction in the oil and gas pools. Not only can it make secondary alteration of hydrocarbons, but generate H₂S and CO₂ gases which can improve reservoir property. The thesis presents two cases in which H₂S is originated from thermochemical sulfate reduction (TSR) and bacterial sulfate reduction (BSR), respectively. H₂S and CO₂ are found in elevated concentrations in Paleozoic reservoirs in Central Tarim.  We have carried out analyses on gas, mineral cement and bulk rock compositions and isotope ratios together with petrography and fluid inclusion to assess the origin of the H₂S.  A deep crustal (e.g. volcanic) origin of the H₂S and CO₂ is unlikely since the inert gases, N₂ and He, have isotope ratios totally uncharacteristic of this source. Considering the starting depth for anhydrite replacement by calcite, homogenization temperatures of inclusions in calcite and burial and thermal history data, anhydrite replacement by calcite is considered to take place in temperatures >125ºC. The temperature is too high for sulfate reducing bacteria to have survived, thus BSR is unlikely. TSR of petroleum fluids by anhydrite in Lower Ordovician and Cambrian carbonate reservoirs is the most likely source of both the H₂S and the CO₂, causing isotopically characteristic pyrite, CO₂ gas and calcite cement. H₂S, and possibly CO₂, migrated into Silurian sandstone reservoirs by cross formational flow. The H₂S, with the similar sulphur isotope ratio as Ordovician anhydrite, was partially lost from the fluid phase by extensive growth of late diagenetic pyrite.  Similarly the CO₂ was partially lost from the fluid phase by precipitation of late diagenetic calcite. 

BSR is more commonly found in the nature than TSR. Sulfate-reducing bacteria (SRB) are generally considered to utilize OAA, alcohol as carbon and energy source but not deplete hydrocarbon directly. To the best of my knowledge, only in one case, analysis of samples from ODP in 1999 showed that SRB may have depleted methane directly. The study on the Hetianhe Field shows that SRB can utilize both OAA and light hydrocarbon gases (LHG), selectively deplete ¹²C, resulting in heavier d¹³C of residual LHG. In detail, H₂S and CO₂ are found in elevated concentrations (up to 2000ppm for H₂S) in the reservoirs near the Carboniferous - Ordovician unconformity in the Hetianhe Field. It is unlikely that H₂S and CO₂ had a mantle component since associated helium has an isotope ratio totally uncharacteristic of this source. Instead, H₂S and CO₂ are probably the result of sulfate reduction of the LHG. Increasing H₂S concentrations and CO₂/(CO₂ + SC_{1- 4}) values to the west of the Hetianhe Field occur commensurately with increasingly heavy hydrocarbon gas d¹³C values from –38.6‰ to –35.6‰. However, TSR is unlikely because the temperatures of the reservoirs ranging from 60 to 80°C are too low, no H₂S or rare pyrite was detected in deeper reservoirs (where more TSR should have occurred) and inferred d³⁴S values of H₂S (from late-stage pyrite in the Carboniferous and Ordovician reservoirs) are as low as –24.9‰. Such low d³⁴S values discount the decomposition of organic matter as a major source of H₂S since organic matter has been measured to have d³⁴S values heavier than -17‰. Bacterial sulfate reduction of the light hydrocarbon gases in the reservoir, possibly coupled indirectly with the consumption of organic acids and anions is most likely. The result is the preferential oxidation of ¹²C-rich alkanes (due to the kinetic isotope effect) and decreasing concentration of organic acids and anions. Relative to ethane and propane, SRB preferentially deplete methane, resulting in a greater positive shift in d¹³C of methane than of other heavier alkane gases. Since SRB are unlikely to have survived since the deposition of Carboniferous and Ordovician, it is important to determine where SRB come. Modern formation water stable isotope data reveal that it is possible that sulfate-reducing bacteria were introduced into the reservoir by an influx of meteoric water from the west by way of an inversion-related unconformity. This may account for the apparently stronger influence of bacterial sulfate reduction to the west of the Hetianhe Field, the greatest increase in d¹³C values of the alkane gases. Thus, empirical equation of methane d¹³C – source rock R_o does not apply in the Hetianhe Field. 

　回主页