Groundwater Pumping Background | A Pragmatist's Take | Douglas Moran | Palo Alto Online |

Local Blogs

By Douglas Moran

Groundwater Pumping Background

Uploaded: Mar 1, 2017

Groundwater pumping (dewatering) regulations are on the City Council agenda for a special meeting on Tuesday March 7 beginning at 6pm. There are two parts: incremental changes for the 2017-8 and to provide Staff with directions for further enhancements,(foot#1) These directions that will determine the options that Staff subsequently presents, which will limit what Council can approve.(foot#2)

There are many conflicting attitudes toward this groundwater at City Hall. For example, in the City's document on water resources, this water is labeled as a "nuisance" and stated that current regulations were the result of "recent public concern over the appearance of wasted water" (emphasis added).(foot#3) The comments to the news coverage over the years demonstrates many misconceptions.(foot#4) My intention here is to provide some background that addresses the biggest of these misconceptions and hopefully reduce some of the clutter in the discussion of the policy.

Recognize that there are two parts to the policy discussions. First, the immediate impacts of dewatering at individual sites involve water and soils near the surface and in the immediate vicinity of the site (hundreds of feet?). Second, the cumulative long-term impacts that affect a much larger area, and consequently involve more factors and variability.

----Intuition on basement dewatering----

The advocates for stronger measures to limit dewatering cite several examples, with the largest being at 736 Garland Drive (map). The Staff report estimates 30.8 million gallons of ground water were pumped (note: earlier Staff reports had different estimates). Part of the advocates' argument is that this pumping is an exorbitant share of a public/community resource. In 2016, 140 million gallons of groundwater was pumped out during the construction of 8 residential basements for an average of 17.5 million gallons. This total is equivalent to the annual usage of 5,000 average homes for landscape irrigation alone (28,000 gallons each) or the total water usage of 1500 homes (roughly 93,000 gallons per home). Using the Garland property for another visualization, the 95 acre-feet of water pumped out during basement construction would be enough to cover that 0.24 acre property to a depth of almost 400 feet. Another visualization is to consider it from a deposit-withdrawal perspective: A very simplistic calculation has this being over 300 years of rainfall on that property.(foot#5) For reference, the expedition of Gaspar de Portola arrived less than 250 years ago (1769).

----"The Aquifers"----

References to an "aquifer" means very different things in different contexts. Different types of soils can hold very different amounts of water and water moves through them at different rates. For example, consider the case of toxic chemicals that seeped into the groundwater at an industrial site. Someone worried about the spread of those chemicals and about protecting drinking water that is drawn from deep underground will talk about the underground strata as forming a collection of aquifers and will focus on where the pollution can most easily move between the aquifers, and will regard layers of clay soil as impervious. On the other hand, someone focused on recharging the deeper stratum understands that even though water moves only very slowly through those clay layers, the massive predominance of those clay layers means that they are major conveyors of water to deeper and deeper layers. They can be more important than the limited locations where the soils are such that they allow faster movement between the various depths.

Note: In many informal discussions, such as here, the terms "shallow aquifer", "water table", and "ground water" are often used in an overlapping manner. I am going to be sloppy in my terminology in service of giving the reader a better intuition about the situations.

----Alluvial fans (deltas)----

One common misconception is that the shallow aquifer is analogous to a lake. Instead, there is considerable variability and complexity to the aquifer. Start with almost all of Palo Alto being built on the alluvial fans of San Francisquito Creek (map) and the other creeks, starting a bit below Foothill Expressway/Junipero Serra Blvd.(foot#6) This is relevant because when a creek overflows its banks (floods), it drops heavier sediment closer to the channel and carries lighter and finer sediments further away. These patterns influence where there are soils that can hold water and how quickly water moves through those soils.

Detail: The banks of a creek get built up incrementally by successive flooding events, especially lesser floods where the small, slower flows drop most of their material near the creeks. With the creek channel becoming higher than the surrounding land there will eventually be a break-out that cuts a new channel. This is the same hydrology as for big rivers, but before I saw the above map, I hadn't thought of its relevance to the local creeks.

Detail: There is a lesser effect for tiny creeks such as Barron Creek. On old maps, creeks like it were simply identified as "Dry Creek", that is, a creek that has water only when it is raining. The old maps show a short fan of oak trees below the Gunn HS fields, with the strong inference that it didn't have a permanent channel to the Bay before the late 1800s. In very large storms, water from Barron Creek cut across the fields to empty into Matadero Creek (roughly following current day La Para Ave then La Donna and Whitsell: map).

Detail: Invisible creeks: Most Palo Altans think that there are no creeks between San Francisquito Creek and Matadero, but when you stop to think about the topology you should realize that there must have been some small creeks that have been supplanted by the storm drain system. Look at the map of the Matadero Creek watershed--I suggest the enlarged view. It shows Santa Rita Creek that may have one reached Bowden Park (Alma and Oregon). Similarly for an unnamed creek that is shown as once having ended on the Stanford campus somewhere near Campus Drive and O'Connor Lane (Koret Pavilion)--perhaps the irregular path there is testimony to its former presence.

Summary: It is easy to fall into the trap of thinking that the creeks have always been where there currently are, and thus you can make predictions about the subsurface soils based upon that. That also applies to creeks that have been re-routed and to those that no longer exist--replaced by a storm drain system. The complexity of what materials got laid down when, where and to what thickness means that the situation in different sections of the city can be very different.

----Earthquake faults----

The next layer of complexity comes from earthquake faults, with most being thought to no longer be active being irrelevant. What is relevant is how they deformed the subsurface materials. Most of the faults under Palo Alto are thought to be various types of thrust faults, unlike the San Andreas which is a strike-slip fault. Some thrust faults involve one side pushing up over the over the other, so that a stratum that is 10 feet deep on one size can be 100 feet deep on the other. Others don't produce much vertical displacement, but the ramming together of the two sides so compacts the soil that it becomes a dam within the aquifer, blocking water movement between the sides.

I learned about these complexities in the 1990s because of groundwater pollution from various sites in the Stanford Research Park moving into my (Barron Park) neighborhood. I am computer scientist, not a geologist. The contractors attempting to plot the spread of the pollutants were surprised by the pattern. First, they discovered that there was a fault under the Bol Park Bike and Pedestrian Path that pushed down the stratum for the shallow aquifer from the Research Park, disconnecting it from the shallow aquifer under most of Barron Park. Unfortunately, this was negated by Matadero Creek. The pollution was in groundwater so close to the surface that it flowed into the creek, and then down the creek past this fault and then percolating into separate shallow aquifers there.(foot#7)

Recognize that maps of faults in this area involve a lot of deduction from sparse data--by the time geologists had the funding to explore for lesser faults, the area was already so heavily developed that it was too difficult and expensive to do so. For example, in the 1.2 mile stretch between Foothill Expressway and El Camino near me, all the maps of the "Quaternary Faults" (active within the last 1.6 million years) list the Pulgas, Hanover and Stanford faults, but some include one or two additional faults and the line on the map for the middle of the Stanford fault zone varying by 0.2 miles (the fine print warns you of this uncertainty).

The maps show no faults between El Camino and the bay, but it is unlikely that this area doesn't have the same patterns found on both sides (the mapping of faulting under the bay is relatively recent). In other words, the gap on the maps is most likely a gap in knowledge, or more famously "The absence of evidence should not be regarded as evidence of absence."


One of the repeated misconceptions in the earlier discussions was that there was a massive sheet of ground water moving down from the foothills into the shallow aquifer on the valley floor. This is a common explanation for the artesian wells that use to dot the valley to the south of here. (Aside: Most of these wells no longer flow because of excessive groundwater pumping.) The analyses I have seen claim that the shallow aquifers under Palo Alto are not connected to the foothills, but rather start at the Pulgas Fault (roughly Foothill Expressway and Junipero Serra Blvd), or the Hanover Fault (Hanover Street extending into the Stanford Campus and Barron Park). Consequently, recharging of the shallow aquifer comes from two sources: local rainfall and percolation from the creeks.

Local rainfall has become less effective in recharging the groundwater because of impermeable surfaces such as buildings, streets, parking lots, driveways, walkways, and patios. I have seen estimates that 20-30% of Palo Alto is covered by impermeable surfaces, but it is hard to do a sanity check: I have visited houses that have only a tiny amount of lawn and landscaping in the front and the backyard covered by a patio, and I have visited houses with extensive yards and gardens. Then there are the parks, playgrounds ... Although the building codes have been updated to include measures to reduce runoff, when a property is predominantly impermeable surfaces, the remaining portions can absorb only so much at a time.

One guess is that between impermeable surfaces and natural runoff, about 50% of rainfall gets into the aquifer.

While San Francisquito Creek still has a natural channel that allows percolation, the other creeks have concrete channels starting at El Camino, or above. Adding water to the creeks during the dry season has to be carefully considered because of how it changes the habitat--both causing problems for native flora and fauna and making it viable for invasive species.


Another misconception from the earlier discussions was that landscaping, including trees, doesn't have roots that go deep enough to reach usable groundwater. Nonsense. Near the bay (Louis Road), the water table starts the dry season around 4-5 feet deep and recedes to 7 feet in summer. In other areas, the water table tends to be found at 10-20 feet deep. Coastal Live Oaks have roots that can go down 36 feet (US Forest Service) in addition to their roots near the surface. If you have a yard, you likely know that an acorn focuses its energy on quickly putting down a long taproot. Other California native plants have very deep root systems, for example, California Purple Needle Grass (still the official state grass) puts roots down 20 feet. And a fair number of non-natives have deep roots.


For the typical soils here (clay), subsidence is not a temporary problem: Once the water has been sucked out, it doesn't readily return. Alviso is the local cautionary example: Excessive pumping in the early 1900s caused it to sink 13 feet. Between the 1920s and 1960s, groundwater pumping caused 2-4 feet of subsidence in Palo Alto, leading to it (and nearby cities) to restrict such pumping.

At the Garland dewatering site, properties 3-4 lots away (220 feet) saw the water table lowered by 3 feet. This amount can cause 0.5 to 1 inch of permanent subsidence. Because of the rate of dewatering, subsidence may not be uniform across those neighboring properties, resulting in enough warping of the frames of doors and windows to cause them to stick or jam, or even in damage to the foundation of the house.

One of the frustrations of the neighbors of these dewatering sites has been that their reports of subsidence have been dismissed with a variety of arguments:
- it is all in the homeowner's imagination,
- it is from other causes, such as the drought (although not explaining why this would affect those houses and not others),
- groundwater doesn't move freely enough to have that effect on neighboring properties, and
- groundwater moves freely enough so that any flowing from the neighbors' properties would be replace by groundwater from beyond that.

----Goes to the bay----

One of the absurdist arguments made repeatedly during previous discussions was that since the groundwater eventually makes it way to the bay, there is no difference in pumping it into the storm drains because it winds up in the same location. This is analogous to a stranger walking into your house and pouring all your wine down the toilet, saying that it was going to wind up there eventually. Or if the City decided that water mains and sewer lines were a nuisance and that water should sent directly from the Hetch Hetchy system to the bay.

----Basements as dams----

Often intertwined with concerns about groundwater depletion by dewatering is a concern that too many basements will disrupt the flow within the shallow aquifer. Recognize that many basements occupy a substantial portion of the property, especially the width (potentially being built up to the minimum setbacks). For example, on a 60 foot wide lot (current minimum) with 6-foot side setbacks, the basement would occupy 80% of the lot width. Actually it is worse than that because certain features, such as light wells, can be located in the setbacks.

The cumulative effects of basements on multiple properties in a small area could be substantial. The problem with predicting this is that we know so very little about the detailed structure of the aquifers. Recognize that groundwater moves easiest through coarse materials--gravel, sand ...--which are also the materials that settle out first (closest to the creek channel). But predicting these deposits requires knowledge about the pattern and size of ancient flooding, and where the creek channel was at that time.

----Rights and wrongs----

Dewatering is an instance of the classic arguments about the tradeoff between individual rights and public/community rights. These well-fleshed out arguments can be useful in thinking about this particular case.

In economics, these discussion are often under the topic of "Externality". For example, if an individual or city chooses to save money by dumping untreated sewage into a river, they are inflicting the costs of sewage treatment on an external party, that is, those downstream who now have to spend more on treating water from the river and who suffer the loss of the use of the river itself for various purposes (example, fishing, swimming).(foot#8)

Another analogy from economics is the thought experiment "The Tragedy of the Commons" in which an unregulated shared resource is over-exploited by individuals, resulting in the destruction of that resource. This analysis is often used to argue for private ownership by creating in a false dichotomy between that and an unregulated shared resources. However, in the real world, the commons of the story were in fact regulated by a mix of informal and formal means and didn't suffer from the hypothetical tragedy.

1. Council Agenda with link to Staff Report (7 page overview with 410 pages of attachments).

2. From my introduction as a blogger: "Real power doesn't reside with those who make the final decision, but with those who decide what qualifies as the viable choices. I stumbled across this insight as a teenager (in the 1960s)."

3. Section ==I "7. Nuisance Groundwater from Basement Dewatering"== in "2017 Water Integrated Resource Plan", tentatively item #3 on City Council Agenda for 2017-03-06 as part of the Consent Calendar (that is, scheduled to be adopted without debate).

4.Stories Chronologically.
News: "Basement groundwater pumping raises concerns", 2010-11-19
"Guest Opinion: Groundwater in Palo Alto is a valuable resource, not construction waste", 2015-12-11
Town Square Forums: "Why is Ground Water Pumping Still Allowed in Palo Alto?", 2015-04-09
News: "Groundwater pumping irks Palo Alto residents", 2015-04-28
News: "Palo Alto residents pumped up about groundwater 'waste'", 2015-11-27
News: "Palo Alto prepares to plumb the mysteries of groundwater", 2015-12-02
News: "Palo Alto firms up rules for pumping groundwater", 2016-02-02
News: "This year, Palo Alto wasted 140 million gallons of groundwater", 2016-12-14

5. 300 years of rainfall has two problems. First it is measure based on the faulty assumption that all rainfall went into the shallow aquifer--no runoff, evaporation or use by plants--and that it stays there, not moving into deeper aquifers or to creeks or to the bay. Second, it uses the recent average annual rainfall of 15 inches because the longer term average is unknown. Aside: After the flooding and the drought of the 1980s, climatologists reported that the previous century had been one of the mildest in California history, and that the norm was big storms and long droughts.

6. Creek & Watershed Map of Palo Alto & Vicinity Oakland Museum of California, 2005.
There is only basic detail on creeks other than San Francisquito.

7. Creeks have gaining and losing segments, that is, segments where groundwater enters the creek because the water table is higher than the creekbed and segments where water from the creek percolates into the ground because the water table is lower. In this area, there can be frequent alternation between gaining and losing segments, and this provides hints about locations of ancient faults that folded the underground strata.

8. I was in college in Boston in the early 1970s. To impress upon us how polluted the Charles River was, we were told (urban legend?) of a student on a LSD trip who jumped into the river. In treating him, the hospital's first concern was the water he swallowed, the second was the water on his skin, then the injuries from the fall.
Similarly, showering was the first priority for members of the crew team (rowing) at the end of practice.

An abbreviated index by topic and chronologically is available.

----Boilerplate on Commenting----
The Guidelines for comments on this blog are different from those on Town Square Forums. I am attempting to foster more civility and substantive comments by deleting violations of the guidelines.

I am particularly strict about misrepresenting what others have said (me or other commenters). If I judge your comment as likely to provoke a response of "That is not what was said", do not be surprised to have it deleted. My primary goal is to avoid unnecessary and undesirable back-and-forth, but such misrepresentations also indicate that the author is unwilling/unable to participate in a meaningful, respectful conversation on the topic.

If you behave like a Troll, do not waste your time protesting when you get treated like one.