NON-HYDROSTATIC GROUNDWATER
Introduction
Non-hydrostatic groundwater differs from normal groundwater in many fundamental ways. First, the forces that cause non-hydrostatic groundwater to flow are not the usual, gravitational body forces that create lateral differences in head; instead, the associated pore pressures are much higher than those within columns of normal groundwater at equal depth. These pressures are accordingly called “anomalously high pore fluid pressures”, or AHPs. Second, the permeability of non-hydrostatic groundwater systems is much lower than that of normal aquifers, which is related to the AHP phenomenon itself. In particular, unlike typical groundwater systems, the weight of the rock column is partly supported by the fluid. Third, most non-hydrostatic groundwater is not potable because it typically has high concentrations of dissolved solids, sometimes accompanied by deleterious concentrations of organic compounds and heavy metals. Fourth, most non-hydrostatic groundwater has resided in the subsurface for a very long time compared to normal groundwater, which is partly responsible for its unpotable condition. Finally, non-hydrostatic groundwaters play important roles in tectonics, diapirism, hydrothermal ore deposition, sediment diagenesis, and petroleum migration, and the associated AHPs can cause severe problems for commercial oil production. For all these reasons, non-hydrostatic groundwaters and their associated AHPs have great scientific and economic importance, in ways that are markedly different than for normal groundwaters. The study of non-hydrostatic groundwater is accordingly challenging but highly informative and rewarding.
This chapter will first review the nature of the forces that generate AHPs and induce non-hydrostatic groundwater to flow, and discuss a few of the consequences. That material will be followed by a brief discussion of the chemical nature of non-hydrostatic groundwater. Finally, descriptions of non-hydrostatic groundwater in several real geologic settings is presented.
Introduction to Anonymously High Fluid Pressures
As one descends into an open body of standing water, he observes that the hydrostatic pressure increases at a linear rate of 0.1 bar/m from the value of one atmosphere at the surface, the latter value representing the pressure of overlying air. The same effect is observed in shallow, unconfined aquifers, with the water table serving as the effective, open water surface. Small deviations from this rate can occur if the water is moving upward or downward,
corresponding to differences in hydraulic head, but those differences are normally small. However, a very different “lithostatic” pressure occurs between the contact points of the mineral grains, and this pressure increases downward at about 0.25 bar/m, because the rock matrix must support its own weight, and normal rocks are about 2.5 times denser than water. The above combination is the “bucket of marbles” scenario discussed in the Groundwater chapter (see GW Fig. 1), and it is normally accurate in the shallow subsurface, where the pore fluids can communicate freely with the water table (Figure 1).
The situation is different at great depth, where the pore fluids are isolated from the remote water table. Under this condition, the pore fluids must also bear the weight of the overlying, saturated rock column, so the pore pressure is the same as the lithostatic pressure. This pressure is in great excess of what would be observed in an open, water-filled, vertical pipe of equal depth, in which pressure would increase downward at the normal hydrostatic gradient. Situations where pore pressures exceed normal hydrostatic pressure are called “AHP’s”, for anomalously-high pore fluid pressures.
The transition zone between normal, hydrostatic pore pressures and AHPs can be rather sharp, and where this occurs at shallow depth, gives rise to many interesting phenomena. For example, AHPs have great practical importance to the petroleum industry. They are mostly responsible for the geological movement and concentration of petroleum into structural and stratigraphic traps, from source rocks in which it was originally present as disseminated, minute blebs. AHPs are also responsible for a drilling problem called “lost circulation”, and in extreme situations for oil well blowouts, which occur when AHPs are suddenly and unexpectedly encountered during drilling and cannot be controlled.
The significance of AHPs extends to many other geological situations, with equally dramatic consequences whose real cause is oft unrecognized. AHPs are probably essential to the generation of conditions where rocks flow or behave plastically. Phenomena such as clastic dikes, diapirs, folding, dissolution and precipitation, and structural creep may be primarily due to, or at least promoted, by AHPs (Figure 2). Low angle thrust faults can move thick rock sequences for tens or hundreds of miles, an unusual phenomenon that is best explained by downslope, gravitational gliding of the rock package over a thin, fluid film under AHP conditions. Much rock deformation and faulting may be due not to brittle failure, as commonly perceived, but as gentle aseismic deformation and creep, promoted by AHPs.
Figure 2. Dike of gray shale cutting the Devonian Huron Shale near Milan, Ohio. The material in the clastic dike was obviously fluidized, probably by AHPs, and plastically flowed into overlying sediments at some stage during diagenesis. Photo from Criss et al. (1988).
Geochemical Character of Non-hydrostatic Groundwater
Basinal fluids, or formation waters, are fluids that reside at moderate depths in sedimentary basins, sometimes under AHP conditions. These fluids are rather rare at the Earth’s surface, and with a few extraordinary exceptions (see below), are mostly encountered in wells drilled by the petroleum industry. These fluids were once commonly called “connate” waters, a term suggesting their derivation from water, usually seawater, trapped long ago during the deposition of their enclosing sediments. Subsequent studies of oxygen and hydrogen isotopes in these fluids reveal that this view is much too simple, and that formation waters generally include a significant component of local meteoric waters.
Most formation waters have near-neutral pH. They also have had a long residence time in the subsurface, at moderately elevated temperatures that commonly reach 150°C or more. Under those conditions, reactions between the fluids and the proximal rock matrix occur, a phenomenon called fluid-rock interaction. These interactions are partly responsible for the distinctive stable isotopic character of formation waters, and more importantly, gave rise to their rather high concentrations of dissolved solids.
The dominant cations comprising the dissolved solids in formation fluids are generally Na and Ca, while the dominant anion is Cl. Chloride behaves conservatively in many systems and therefore serves as a useful reference ion. These three ions greatly dominate the dissolved load in typical formation waters, so their description is of particular importance.
Davisson and Criss devised a straightforward method to compare the relative concentrations of these ions and to provide insight on the geological processes that generated them. This method compares the Na and Ca concentrations of a particular sample of formation water to those that would be observed in seawater that was diluted (or enriched) to the same chlorinity as the sample. The deviations of the sample concentrations from these latter reference values, expressed in equivalents to facilitate interpretations that incorporate charge balance, are called the “Na-deficit” and the “Ca-excess”, simply because the Na concentrations of formation waters generally lie below the seawater dilution line, while the Ca concentrations generally lie above it. Those quantities are calculated as follows:
where the concentrations observed in the sample (subscript meas) and in average seawater (sw) are all reported in mg/l, while the numerical multipliers on the right convert the results to milliequivalents per liter.
Remarkably, when the above relationships are used to analyze samples of formation waters from all over the world, hosted in rocks ranging from carbonates to shales to granites, a linear relationship with a small y-intercept and a slope very close to unity is found (Fig. 3). Davisson and Criss called this relationship the “basinal fluid line”, or BFL, and explained that the unit slope of this trend proves that exchange of two Na ions for each Ca ion, were involved in the formative process of these fluids (Fig. 3).
The global coherence of the BFL indicates that exchange between heated fluids and feldspar, particularly plagioclase, must be involved in the genesis of most formation fluids. This deduction is very reasonable, because feldspar, taken collectively as a family, is the most common mineral in the Earth’s crust.
Other geochemical processes such as seawater evaporation will produce fluids with different ionic characteristics that define trends on the Ca-excess vs. Na deficit plot that are distinct from the BFL. Such fluids are much less common than those conforming to the BFL, but in any case the “excess-deficit” plot provides a useful way to elucidate their character and identify probable formation mechanisms.
How do crustal fluids get in sedimentary basins? When a sedimentary basin such as the Mississippi Delta is forming, pore fluids are included in the deposits. As additional layers of sediment are added, the weight of those superjacent layers compresses those below, reducing their permeability. Eventually, when depths commonly between one and two miles are attained, the pore fluids become isolated from the surface, and must begin to shoulder the weight of the overlying column of saturated rock. As soon as these fluids carry some of the rock load, pressures rise above those in a hydrostatic column, so the AHP regime is entered. Various reactions including clay dehydration can contribute additional fluid, while magnifying the AHPs. Movement of these fluids can help oil droplets that are maturing in the same sediments to migrate upward, or in rare instances, to move into economically important traps. Of course, burial of these sediments to such great depths requires 10 million years or more, giving the pore fluids plenty of time to reach with the enclosing sediments and acquire their distinctive ionic concentrations.
AHPs occur worldwide, but their recognition has been primarily due to petroleum exploration, so their real distribution and consequences are likely underappreciated. Because observational experience has been primarily linked to oil, AHPs are mostly known in Tertiary sediments and convergent margins, particularly in areas of sedimentation, burial compaction, diagenesis, and tectonic compression.
Gulf of Mexico
In the USA, AHPs are best known in the Gulf of Mexico, particularly in the offshore parts of the Mississippi delta. Sedimentation coincident with burial compaction has been very active and nearly continuous in this region since the Eocene, so a 10-20 km-thick wedge of clastic Tertiary deposits has accumulated. As the sediments are compressed by the weight of younger, overlying deposits, pore space is lost and dewatering occurs. At the same time, temperatures increase, trapped organic matter begins to slow transform into oil, and diagenetic mineral transformations occur in what is actually a process of low-grade hydrothermal metamorphism. Burial compaction of itself reduces permeability and porosity, progressively. The important digenetic transformation of montmorillonite to illitic and chloritic clays also involves loss of water, so it is essentially a dehydration reaction that releases water into sediments that are already well compacted, also promoting the generation of AHPs deep in the sedimentary wedge. All of these transformative processes are linked to the generation, maturation, and migration of oil, so they are of great importance to industry. The AHPs also cause problems when encountered during drilling into these environments, as for example occurred in the Deepwater Horizon blowout disaster of 2011 (Figure 4).
Figure 4. The Deepwater Horizon blowout of April 2011 resulted caused 11 fatalities and despoiled the Gulf of Mexico with 4.9 million barrels of oil. The blowout occurred when AHPs were encountered at an extreme depth of 18400 feet, representing the combination of the water depth of 5070 feet and 13,300 feet of rock. Photo by US Coast Guard.
Examples of anonymously fluid pressures are abundant, most examples from the petroleum industry. Typically, what is observed is pressure will initially follow a hydrostatic curve and at some point, will veer off the hydrostatic curve and migrate towards the lithostatic curve of Figure 1. Sometimes with increasing depth the pressure will migrate back to the towards the hydrostatic curve, due to some pocket of pore fluid under high pressure. With increasing depth, the pore fluid pressure will obtain lithostatic pressure due to the overburden of the rock layers.
Convergent Margins
Convergent margins are also a place where AHPs are observed. Convergent margins are areas where tectonic plates meet in a compressional stress relationship. Subduction zones, where ocean plate subducts under continental plate due to its lower density. Along that margin where subduction takes place an accretionary wedge of sediment from the ocean floor accumulates and is focal point of AHPs. Many times there are fluid emanations from the accretionary prism with exotic chemistry. These accretionary prisms are under water sometimes many miles below the ocean surface. In Figure 5 is a Google Earth image of the Barbados convergent margin where the Atlantic plate subducts under the Caribbean plate and builds up an accretionary prism.Figure 5. Accretionary prisms are area with AHPs, where sediment is caught between the continental margin and subducting ocean plate. The sediments are either pelagic sediments scraped off ocean floor or clastic material washed off the island arc.
Mud volcanoes occur throughout the world and are also associated with AHPs. They typically occur in oil and gas rich basins, or convergent margins. The mud volcano is a manifestation of diapirism and heating of subsurface. The mud exploits faults and fissures as it ascends. A mystery is how the mud formed from solid rock. A famous example occurred in 2006 when a drill crew working east Java, Indonesia was exploring for oil, and punched into a steam zone, causing steam and then mud to rise up to the surface. The peak flow was 180,000 m3 per day, and mud is still flowing. Most mud volcanoes are associated with natural gas or methane. Several natural mud volcanoes occur in northern Mendocino County, California that mostly emit gas dominated by CO2.
Rumsey Hills, California The Rumsey Hills provides a remarkable and informative instance where formation fluids from an overpressured reservoir can be collected as saline springs emerging at the Earth’s surface. These hills are situated on the eastern most flank of the Coast Ranges, immediately west of California’s Great Central Valley, and attain elevations in excess of 1600 feet. Uplift of the hills probably started during the Quaternary and continues today. The deformed clastic sediments are part of the Great Valley Sequence, mostly of Cretaceous age. An extinct accretionary prism called the Franciscan Formation makes up the core of the California Coast Ranges. This famous formation includes discontinuous blocks of rock that don’t correlate, much like actively deforming sediments in modern accretionary prism.
The San Andreas transform fault system is part of a regional tectonic regime that places the entire region under a state of compression. This compression drives mountain building, and generates and maintains AHPs. These AHPs are known from exploratory oil wells, and their impact is obvious to the field geologist. Literally hundreds of springs, whose flows are all driven AHPs, occur in the California Coast Ranges, and dozens of these are in the Rumsey Hills (Fig. 6).
Figure 6. Saline spring emerging at high elevation in the Rumsey Hills, being investigated by Lee Davisson. Note the white halite deposits and the salt grass. Expulsion of these formation fluids is occurring at high elevation, a situation completely at odds with how normal groundwater would flow. It is indeed this compressional tectonic drive and the resultant AHP condition that makes these fluids accessible to the field geologist. Photo by REC. The Rumsey Hills is part of this compressional regime. These hills are constructed of recumbently folded, upper Cretaceous clastic sediments that have been thrust faulted, and overlie a thick, east-tapering wedge of Cretaceous sediments that is being driven, or underthrust, beneath the hills (Figure 7). Figure 7. Structural cross section, looking north, of the Rumsey Hills, showing recumbent fold and thrust faults. After Davisson et al., 1994. This structural response is clearly due to active tectonic shortening and compression of this thick package of sediments. A measurement of the rate of dewatering was conducted in Rumsey Hills on the east side the Coast Ranges, and it was one liter per second, which representing the volume strain rate of this region. Geothermometers were used to estimate source temperatures for each measured spring, and it turns out the deepest spring waters formerly had temperatures of about 90ᵒC, which places their source area about 4 km below the surface, in the core of the recumbent fold. Figure 8. The relationship between close in temperatures and depth in the Tippets well, drilled in the Rumsey Hills in the 1950s. At a depth of 4 km depth, the close in temperature corresponds to our geochemically estimated source temperature of the spring waters. Note the gradient down to 4 km in depth is 11℃/km, this due water moving up muting the actual gradient. Interestingly, while the remarkable AHP condition in the Rumsey Hills is created by tectonic forces that operate on a very large scale, the AHPs themselves control much of the local tectonic response. In particular, and contrary to much geophysical thinking, the AHPs promote a plastic, rather than a brittle, response of the host rocks. Melchiorre et al. demonstrated that earthquakes are effectively quenched where the AHPs are present, even though faulting, diapirism, uplift and other and tectonic displacements are actively occurring. Clearly, the presence of the warm, overpressured fluids promotes dissolution, recrystallization, tectonic creep, and other responses that are much more gentle than those that occur in brittle rocks. A simple analogy is provided by the bending of two sticks, one wet and green, and one dry and brittle. The chemistry of the Rumsey Hills springs is remarkably similar to formation waters associated with oil and gas fields. These saline springs conform to the BFL, are up to 2/3 as salty as seawater, include a meteoric water component, and have reservoir temperatures of up to about 100°C. Additional details are provided by Davisson et al.