Groundwater separated from atmospheric pressure by relatively impermeable material is termed confined groundwater.  When such zones are penetrated by wells, the water rises above the point at which it was first found because a confined aquifer is under pressure exceeding that of atmospheric pressure.  Confining beds vary in permeability and, hence, in their ability to confine artesian aquifers.

A major difference from the unconfined aquifer is that when an artesian aquifer is pumped, there is no dewatering of the saturated zone by gravity discharge.  A well that taps an unconfined aquifer above a confined aquifer can dewater the former by gravity drainage and not affect the artesian aquifer if the confining bed between them has negligible permeability.

The potentiometric surface is an imaginary surface above the aquifer, to which water from an artesian aquifer would rise in a pipe.  The term potentiometric surface means head- or potential-indicating surface and is preferable to the term piezometric surface, which is found in some literature.  In the early development of some artesian basins, the potentiometric surface was above the land surface giving rise to a flowing artesian well.  More commonly, the potentiometric surface is above the top of the artesian aquifer, but below the land surface.  This type of well is referred to simply as an artesian well.

The release of water from artesian storage differs significantly from the way water is released in an unconfined aquifer.  The best way to visualize the source of water in an artesian aquifer is to consider a typical situation — an artesian aquifer consisting of shale with negligible permeability.  Consider that this sandstone is saturated and overlain by 500 feet of confining beds having an average density of about 2.4 (weight per unit volume = density; density of water = 1).  In this case, the top of the aquifer supports a load of rock equal to about 520 lbs. per sq. in. (36.4 kg/cm²).  A part of this load is supported by the sandstone aquifer and a part by the water, which is under artesian pressure and hence is pushing upward and downward against the confining beds.  When the artesian pressure is reduced, as happens near a discharging well, the ability of the aquifer to support the load of rock is reduced by an amount proportional to the reduction in artesian pressure; as a result, the aquifer collapses a little or is compressed. Some water, which is thereby released from artesian storage, moves toward and out of the well.  Further, at the same time the loss in some of the artesian head permits the water to expand a little and thus releases more water from storage.  The action is much like squeezing a wineskin.

Oscar E. Meinzer was the first to recognize and develop qualitatively this theory of the compressibility and elasticity of artesian aquifers (1928, Compressibility and elasticity of artesian aquifer, Economic Geology 23, no. 3, 263-291).  In 1935, C.V. Theis made the first quantitative determination of the amount of water given up from storage in artesian aquifers.  His example using heat conduction allowed the development of a mathematical theory that led to the equation for the flow of groundwater, through permeable media to a discharging well.  For a complete discussion of this theory and others for determining T (transmissivity) and S (storage) by field flow test, see Theory of Aquifer Tests, USGS Water-Supply Paper 1536-E by J.G. Ferris, D.B. Knowles, R.H. Brown, and R.W. Stallman.

Theis defined the coefficient of storage as "the volume of water, measured in cubic feet, released from storage in each column of the aquifer having a base 1-foot square and a height equal to the thickness of the aquifer, when the water table or other piezometric surface is lowered 1 foot."  It becomes obvious that the coefficient of storage is proportional to the thickness of the artesian aquifer.  For most artesian aquifers, the values range from 10-6 to about 10-3.

Rate of spread of the cone of depression in artesian aquifers

There is a large difference in the rate of spread of the cone of depression (cone of influence) around a discharging well in an artesian groundwater reservoir (1965, S.W. Lohman, Geology and Artesian Water Supply, Grand Junction Area, Colorado, USGS Professional Paper, 109) as opposed to that which develops when a water table aquifer is pumped.  In an unconfined (water table) aquifer, a large volume of water drains slowly by gravity from the sediment within the spreading cone.  In an elastic artesian system, the pressure change traverses the aquifer with the speed of sound; the cone of depression and the area of influence (in which drawdown takes place) grow very rapidly, but at a gradually diminishing rate.  The area of influence of the cone of depression in an artesian aquifer pressure surface is commonly several thousand times larger than that in an unconfined aquifer.

Elastic properties of artesian aquifers

J.G. Ferris of the USGS discussed this effect in an unpublished memorandum (circa 1947) as follows: "It is a common observation that wells in some areas undergo changes in water level during periods of large fluctuations in barometric pressure. … Under the influence of increasing barometric pressure the exposed water level in the well casing is depressed.  An equal effect is transmitted through the soil-air column to the shallow water table with a resultant balance of pressure inside and outside the well tapping the shallow aquifer, and consequently no net change in water level occurs.  In the deeper aquifer where water is confined under pressure, a part of the transmitted load is borne by the solid matrix of the confined aquifer and the balance is borne by the confined water.  Consequently, the water level in the well tapping confined groundwater is depressed an amount equal to the difference between the barometric pressure change and the portion of that change which is borne by the confined water, or not borne by the rock matrix.  The ratio of water-level change to the barometric change, in equivalent units, is termed the barometric efficiency of the aquifer. Note that the effect is inverse, that is, as the barometric pressure raises the water level declines."  Heavier, denser air at a higher barometric pressure pushes harder against the water column.  These differences are not great.  Note that a very low barometric pressure of 28.70 allows the water to rise only less than 1-1/2 feet above the level of 29.80.

Subsidence

Subsidence is the lowering of ground level, and can occur anywhere for a variety of reasons.  It is a particular problem in the American Southwest and regions of the High Plains where overpumping of artesian aquifers is occurring.  This happens because the formations become compacted as the volume of groundwater is withdrawn, leaving empty space that is closed by the settling action of the heavy overburden.

This farmer is looking at land that has subsided in altitude.  This picture shows a small example of what can happen when land loses elevation, sometimes very quickly.  Photo courtesy USGS.

When recharge equals or exceeds withdrawal, the water table will rise to or above its former level after pumping.  However, if more water is withdrawn than can be recharged to the aquifer (overpumping or "mining"), the water table is lowered permanently.

When the water table is lowered for an extended period of time, the pore spaces are emptied of water long enough to permit the replacement of the water by grains of sand, silt, or clay.  Where water is removed from fractures and joints, the weight of the overlying rock material gradually closes these spaces.  Subsidence is greater in formations with high clay content because clay is highly compressible.  Clay holds a great deal of water, as mentioned earlier, but does not allow it to flow very readily.  A large percentage of the volume of some clay beds is actually water.  When the water is removed, the formation collapses.

Once a potential trouble area is identified, steps should be taken to properly manage the region's groundwater supply.  This is indeed being done in many parts of the arid West.  Management can be accomplished by adequate spacing of wells, limiting the amount of withdrawal, and developing systems to encourage recharge.  Since the pumping of sand is another means of removing material from an aquifer, careful analysis of formation composition is important in choosing screens and gravel packs that prevent sand pumping.

If production wells extend below thick formations that have become compacted, the bottom of the casing will remain stationary while the overlying material settles downward.  The resulting stress will usually collapse the casing.  Compaction of the water-producing formation and overlying formations results in permanent reduction of the groundwater storage capacity of a region.  This is the most serious consequence of all, especially since this most often occurs in the arid and semiarid regions that rely most heavily on groundwater supplies.

Recharge of artesian aquifers

In arid regions, outcrops are commonly dry or nearly devoid of moisture most of the time.  This is true for exposures of aquifers as well.  The outcrops of these permeable strata receive water (recharge) from precipitation or from streams that cross the outcrops.  These intake areas are very important for it is here that water enters the aquifers and moves slowly down the dip of the strata beneath the overlying confining beds toward lower areas of natural discharge or artificial discharge, such as through wells.

In dry regions, little water enters the outcrops directly from precipitation because most of the water falls during the hottest part of the day as afternoon thundershowers.  However, rains in the evenings and spring runoff from melting snow in cooler times of the year do contribute some water, which does infiltrate.  Most recharge occurs where the outcrops are crossed by streams or irrigation canals and drainage ditches.  Sometimes interformation leakage occurs from below the confining bed where the aquiclude is more permeable or because part of the confining bed is missing.  Further, some water may leak directly from the confining bed into the artesian aquifer.  In faulted areas, some artesian water from aquifers under relatively high pressure can leak upward along the fracture zones that commonly bound faults and, hence, recharge other artesian aquifers where hydrostatic heads are lower.

The velocity of water moving through a permeable artesian sandstone is very slow.  By using a modification of Darcy's law developed by prominent hydrologist L.K. Wenzel (1942, Methods for Determining Permeability of Water-Bearing Materials, with Special Reference to Discharging-well Methods, USGS Water-Supply Paper 887), the velocity of water in a typical artesian aquifer can be computed as:

v = PI/7.48θ

in which v = velocity, in feet per day; P = coefficient of permeability, in gallons per day per square foot; I = hydraulic gradient, in feet per day per square foot; 7.48 = number of gallons in 1 cu. ft.; and θ = porosity, as a decimal fraction.

This velocity indicates why recharge takes place slowly.  Further, it also indicates why recharge must be considered in the evaluation of the total amount of groundwater available for development.  If more water is discharged by natural plus artificial means than is recharged, the potentiometric surface must drop a little each year and, hence, the cost of pumping groundwater must increase.  If water managers seek to stabilize the potentiometric surface, they must make sure that discharge does not exceed recharge.

One way to increase artificial discharge (through wells) and yet not decrease the total discharge is to intercept natural discharge with wells.  Another important significance of the slow velocity in a typical artesian aquifer is that the quality of the groundwater must change as it infiltrates and traverses the aquifer to its point of discharge.  This groundwater is in the transient state for hundreds, if not thousands of years.  By the same token, the length of time in transit is a factor in removing certain impurities which may be present in the water at the time it infiltrates, and in the degree of mineralization of the groundwater at its point of use.

Fine-grained and nearly impermeable shales and clays (aquicludes) of great thickness and areal extent are common elements of all artesian basins in the United States.  A common misconception is that these fine-grained materials do not leak.  If this were true, our artesian aquifers would contain only the water that was locked up in them originally.  In most cases, this water would reflect the composition of the waters in the environment of deposition, namely the original sea water (connate water).  Further, if the artesian aquifer could not be recharged through the confining beds that overlie or underlie it, there would be no water left in some of the heavily pumped artesian basins of the Great Plains and Rocky Mountain regions.

The easiest way to visualize the effect of slow leakage through shale or clay is to think of the size of the area in which the aquiclude confines the aquifer.  A typical township in the United States consists of 36 square miles and contains more than a billion square feet. Multiplying the number of square feet in one township by the number of townships defining the extent of the artesian basin, one arrives at the total number of square feet delineating the aquiclude.  If only a very small amount of water moves through each square foot of the aquiclude in a day or a month, it is obvious that the sum total of water added to the artesian aquifer this way must be astronomical.

This enables us to conclude that artesian aquifers are not only recharged through their outcrop areas, but also through their confining beds.  The directions taken by the groundwater as it leaks across beds confining the aquifers is dependent on the head relationships above and below the confining beds. The flow of groundwater is always from regions of high head (pressure) to regions of low head.

Discharge of artesian aquifers

Water that moves through artesian aquifers is discharged upward through areas of the overlying confining bed that are relatively permeable.  Fault and fracture zones that penetrate the confining bed would also allow the water to escape upwards.  Over a period of several years, the total rate of natural discharge by upward movement of the groundwater from the artesian system is equal to the aggregate rate of recharge except for the water that is discharged from wells.

In dry areas, the upward escape of water is not noticeable because the rates of evaporation are so high that rock and sediment surfaces are kept dry most of the time.  Further, the amount of water contributed to streams from artesian aquifers in an arid region is so small as to be immeasurable.  In the southwestern coastal basins of California, for instance, some groundwater from artesian aquifers is discharged into the Pacific Ocean below sea level.

In humid areas, a great deal of water is discharged naturally from artesian aquifers to springs and streams.  Seepage into streams and flow from springs form the base flow of many streams in the Atlantic Coastal Plain.

In the Savannah, Georgia, area, some groundwater from the principal artesian aquifer seeps through confining beds into the shallow Quaternary sands at the surface or into the Atlantic Ocean (1963, H.B. Counts and E. Donsky, Saltwater Encroachment Geology and Ground Water Resources of Savannah area, Georgia and South Carolina, USGS Water-Supply Paper 1611).  This takes place where the hydrostatic head in the artesian aquifer is greater than that in the overlying rock units or is above sea level.  The pumping of large quantities of groundwater in the Savannah area has greatly altered the direction and velocity of groundwater movement.  Management will be required to prevent salt water intrusion into fresh aquifers.  If pumping does not exceed surface recharge, infiltration should not occur.

Low-yielding wells in artesian aquifers

On occasion, it is necessary to develop aquifers that have very low yields.  Sandstone aquifers of Paleozoic or Mesozoic age are commonly in this category.  Many areas of the United States have intensely developed artesian basins.  In the early history of these areas, wells yielded water by natural flow.  Wells having sufficiently large artesian heads generally were connected directly to home water systems or to elevated storage tanks.  Some wells having small heads and flows were allowed to flow into cisterns from which water was pumped when needed.  In dry regions such wells were operated at nearly constant drawdowns and as head diminished so did the discharge.  In most cases, the reference points below which the drawdown occurred were the heights above land surface to which the water would rise when the wells were shut-in.  Few wells in the initial development of these artesian basins were shut-in and tested for long enough periods to determine the proper reference point or static level prior to development of the region.  So, in many cases, the maximum amount of drawdown which has occurred since development cannot be computed.

In other parts of the country, particularly the humid East, or Southeast, a well having a specific capacity of less than 0.1 gpm per ft. of drawdown would be considered a dry hole, for many wells have much higher specific capacities, i.e., 10s or even 100s of gpm per ft.  In western Colorado, however, a specific capacity of less than 0.1 indicates a well of great value as the shallow groundwater from alluvium of Holocene age in the stream valleys is of inferior quality and cannot be used.

Commonly, in the sequence of development of artesian basins in the West, wells were allowed to flow for a period of months or years until greater development increased the problems of well interference.  The discharge of one well would cause the artesian head to decline in a nearby well.  Further, declining natural flows, which are proportional to available artesian heads caused discharges to be reduced so that pumps were installed to restore well yields to those of earlier flows.

Noisy wells in artesian aquifers (an extreme example of a common occurrence)

In a confined aquifer, changes in barometric pressure can cause a well to blow or suck by moving updrafts or downdrafts of air through the pipe.  A confining bed with negligible permeability must exist over the aquifer for this to occur.  Under the influence of increasing barometric pressure, the exposed water surface in the well casing is pushed down.  A part of the transmitted load is sent through the water in the aquifer, but part is absorbed by the solid matrix of geologic stratum containing the water.  The ratio of water level change to the barometric change, in equivalent units, is termed the barometric efficiency of the aquifer.  Thus, as pressure goes up, the water level goes down and the well draws or sucks air.  As pressure decreases, the reverse is true and the well releases or blows air.  In some wells, the frequent cloudiness or color in well water preceding a storm can be explained by the rapid rise of the water level that accompanies the decreasing pressure as the barometric low moves in.  As water rushes into the well, it carries fine-grained material, silt, and clay from the formation into the open hole or through the screen.  Further, some gas is released from solution as the pressure decreases and this serves to increase the cloudiness of the water.

It must be noted, however, that noticeable effects on artesian wells from changes in barometric pressure are rare.  Percentage changes in pressure across even the most abrupt weather fronts are small.  In addition, water as a liquid compresses and expands only slightly under changes in pressure, much less than the air.  The effect on most artesian wells represents a "slight sigh" rather than a noticeable "blowing" or "sucking," although this does occur.

The above information is excerpted in large part from Chapter 15 of the 1999 NGWA Press publication, Ground Water Hydrology for Water Well Contractors.