One of the most cussed and discussed concepts is whether there is a billion acre-feet, or two billion, or some other huge volume of groundwater contained in the aquifers of the Valley.  We still see these numbers being cited by some people as justification for new appropriations of confined aquifer water, 50 years after such numbers were first noted by the U.S. Geological Survey (USGS).  Do these numbers have any basis in reality?  Realistic or not, are these numbers meaningful?  

The USGS stated that the “ground water reservoir” of the Valley covers an area of about 2 million acres. This is equal to 3,125 square miles, probably a fair, although rough, estimate of the size of the entire Valley floor in Colorado.  For their estimate, the USGS made an assumption that the entire Valley is underlain by a productive aquifer from the ground surface down to 5,000 feet depth, with a 20% specific yield from top to bottom.  The term “specific yield” refers to the percentage of the aquifer’s total volume that will drain by gravity if it is allowed to do so. These assumed depth and yield numbers are simply not credible, nor are they realistic.  The data we have shows that below about 2,000 to 3,000 feet on average in the northern part of the Valley, and much shallower in the southern part, the sediments become much denser and harder, groundwater yield is lower, and low permeability is common.

To follow the USGS’s thinking in the early 1970’s and arrive at a number of one or two billion acre-feet, you also have to assume that even if you could pump the aquifer down to 5,000 feet economically (a laughably unrealistic assumption) the water would all be of good, potable quality.  The available data, in the form of geophysics and oil and gas test data, show that the water quality is a lot poorer in the “passive” lower-permeability confined aquifer zone.  

Also, if you want to pump all of the groundwater down to a depth of 5,000 feet you would have to de-water (drain by gravity) the aquifer down to that great depth. The physical reality of drawdown cones of depression means that even if a large amount of groundwater exists, of whatever total volume, you physically cannot pump it all.  You cannot do it, even if you construct and pump a 5,000-foot deep well on every section corner in the entire Valley. Think about the nature of well pumping and deep well drawdown:  due to the funnel shape of cones of depression from pumping wells, as you draw the aquifer level down you need to add supplemental wells between your initial grid of wells to keep up the production.  Then, as you draw the aquifer down even further you need to add even more wells between the previous wells at even closer spacing, and so on.  It is a hypothetical race to the bottom with a geometrically increasing number of supplemental wells and pumps, but in reality it simply can’t be done.  

However, let’s do a hypothetical thought experiment.  Let’s assume that you install and pump hundreds and hundreds of 5,000-foot deep wells Valley-wide, and let’s further assume that you succeed in drawing down the water level to 5,000 feet all the way across the entire confined aquifer.  What would be the impact?  Due to the high degree of hydrologic interconnection in the Valley, before you get very far in drawing down the confined aquifer, you will first begin to dry up the unconfined aquifer, and you will severely deplete many of the streams and other surface water features in the early stages of this race to the bottom.

Also, if you pump the confined aquifer hard enough to draw the water level down into the aquifer sediments, you may cause harm in ways totally unrelated to stream depletions or neighboring wells going dry.  You may cause injury in ways we cannot even envision due to the limitations of what we presently know about the hydrologic system.  The USGS, 50+ years ago, did good work in collecting basic data about water levels, water quality, and mapping groundwater flow directions in the Valley.  Unfortunately, the back-of-the-envelope estimates of one or two billion acre-feet were wrongheaded and just plain incorrect.  The USGS never should have published these numbers, because it gave totally unsound notions the imprimatur of USGS approval. Every proposed groundwater export plan for the past 40+ years has cited these misleading USGS numbers as one justification for massive new appropriations from the confined aquifer.  We have been fighting these misleading numbers ever since AWDI, if not before.

But what about the deeper layers of the confined aquifer?  Does the complex aquifer layering mean that we can pump groundwater from the deep “passive” zone of the confined aquifer, and not impact the shallower “active” zone or the unconfined aquifer?  Emphatically, no. If we pump from a deeper zone, whether of lower permeability or not, there are still continuous hydrologic connection pathways throughout the aquifers from bottom to top, and we cannot avoid causing drawdown (i.e., reduction in confining pressure) that would propagate upward and outward through the layers.  

The takeaway is this: it is irrelevant whether there is 1 million,100 million,1 billion, or some other number of acre-feet of groundwater in the Valley’s aquifer system.  It is important to bear in mind that the Valley’s hydrologic system is overappropriated: there are more water rights than there is water physically and economically available to supply them. The hydrologic system, throughout all of its layers and including the surface streams and other surface water features, is interconnected in many areas and in many complex ways.  This means that any new or additional pumping from this overappropriated system must be presumed to cause negative impacts to existing water users unless all potential modes and amounts of injury are fully analyzed, accounted for, and mitigated.  This is one reason why the Valley’s water users and the Colorado Division of Water Resources have gone through the painful, expensive, complicated, multi-year process of crafting and adjudicating rules on groundwater pumping.

Over-pumping of aquifers that are geologically similar to the Valley’s confined aquifer, such as in the San Joaquin Valley of central California, parts of the Phoenix basin of Arizona, the valley of Mexico City, and elsewhere in the world, have resulted in permanent, non-recoverable land subsidence.  Also, increased levels of arsenic in groundwater have been shown to occur during heavy pumping in the San Joaquin Valley.  Arsenic is not uncommon as a naturally occurring mineral in confined aquifer water in the San Luis Valley, sometimes in concentrations that exceed EPA’s drinking water standard.  

Will any of these, or other, impacts happen in the confined aquifer?  Based on what we know now, it is difficult to say whether either injurious land subsidence or water quality changes would occur due to new concentrations of confined aquifer pumping in the Valley.  But we know that geologically similar basins have experienced such impacts due to aquifer over-pumping.

We may be near the edge of our knowledge of how severe the impacts will be that we can predict with confidence, if there is additional pumping of the confined aquifer in the Valley.  Injurious stream depletions. Well to well drawdown interference. Land subsidence. Water quality degradation. All of these impacts have been observed in similar groundwater basins. Will they happen in the San Luis Valley? And if so, will they be so severe as to cause injury?  The RGDSS model, the best and most reliable model of the Valley’s hydrologic system so far constructed, can answer some of these questions, but maybe not all, at least in its current form and with our current knowledge of the hydrologic system.

Scientists and engineers are often asked to predict outcomes where livelihoods and property are at stake. This happens any time an engineering team designs a building, a bridge, or a dam.  It also happens when scientists use their best data, tools, and knowledge to predict the outcomes of pumping an aquifer.

In the past two decades we have learned a tremendous amount about the confined aquifer, as well as the unconfined aquifer and the surface hydrologic system.  There is still much to learn, and thus there may be unintended consequences.  In this author’s view, an open mind, along with a large dose of caution and a healthy skepticism about any new pumping proposals should be the watchwords for the scientists, engineers, managers, and water users struggling to achieve and maintain sustainable water supplies in the Valley.

By Eric J. Harmon, P.E., Hydrogeologist