Well Field Woes

Closed Canal and drought reduce levels in the Great Flats Aquifer

The Great Flats Aquifer is the primary sources of municipal water in the Capital District. The aquifer was originally deposited by an earlier version of the Mohawk River and now it relies on the River for a major fraction of its recharge.  The operation of the locks and dams of the Erie Canal in this area result in a full-pool setting in the navigation season (May to November) where connection with the aquifer and recharge are significant.  However, in the non-navigation season the removable dams are retracted and the pools are drained and groundwater levels fall. 

The 2020 Pandemic has interrupted opening of the canal, and thus we are in a summer season, but in non-navigation conditions, which means reduced aquifer recharge. The problem with the aquifer has been compounded by a precipitation deficit in the last few months. 

Thus rainfall is low, the Mohawk is low, and groundwater levels are low.  As such there are concerns about water resources for the rest of the summer.  One special concern is that municipal wells may end up drawing water from progressively more distant parts of the aquifer, in which case they may tap into groundwater contaminated by Volatile Organic Compounds (VOCs), which are locally problematic because they contaminate parts of the aquifer.

1. Great Flats Aquifer

The Great Flats Aquifer is one of the most productive and reliable aquifers in Upstate New York [1].  It consists of sands and gravels deposited when an earlier version of the Mohawk emptied into Glacial Lake Albany [2]. This setting resulted in the deposition of deltaic deposits that are sand- and gravel-rich, and very transmissive to groundwater flow [3].  While glacial till and finer sand and silt occur in the lower area, the best wells in the aquifer take advantage of the well-sorted sands and gravels in the upper 50-60 feet of the deposits. The permeable transmissive layer is underlain by relatively impermeable glacial till and bedrock.  

Water from several different well fields supply all or most of the water for Glenville, Niskayuna, Rotterdam, Rotterdam Junction, Schenectady, and Scotia.  They indirectly supply water to Ballston Lake and Clifton Park. Thus the Great Flats Aquifer is a critical municipal water supply.

Map of the primary municipal wells in the Great Flats Aquifer (pink, yellow, and orange), one of the most productive aquifers in NY State. Lock 8 and Lock 9 and their associated removable dams are typically in place by the month of May for the navigation season, but not in 2020 due to Covid19.  Thus we do not have the typical pool configuration that is so important for recharge.  Lack of pools combined with dry conditions has resulted in very low levels of groundwater (map modified from George et al., 2014[4]; the mapped extent of the TCE plume from the Scotia Depot is indicated in shades of dark purple [5]; and an estimate of the remediation of contaminated soil at the Schenectady International plant to the NW of the Glenville well field from the NYS DEC [17]).

The gravels are so transmissive in the Great Flats Aquifer that the primary source of water is the adjacent and overlying Mohawk River: the gravels act as a giant filtration system [1,2,3]. We depend on this fragile symbiotic relationship between the River and our municipal water supply.

Movement of water through the gravels is rapid, so what happens in the River can affect water in the wells quickly.  Seasonal variation of river water temperature is large enough and ground water movement is fast enough to measure changes in the temperature of aquifer water due to changes in the river.  Work by Winslow and colleagues established that seasonal temperature changes took about 25 days to reach the wells [2].

Over 50 years ago, radioactive tritium from tests of thermonuclear devices resulted in fallout in the Schenectady area, and significant pulses of radioactive tritium (3H in H2O) could be measured in the river and then later in water of the Schenectady well field [2]. The first serendipitous “experiment” was between April and June of 1953 when weekly samples showed it took about 48 days for the radioactive water to show up in the well.  The second experiment was in 1961 when the initial peak of tritium in river water of the Mohawk occurred on 25 September, and this tritiated water was then measured in the Schenectady wells by about 9 October, a transit time of 18 days [2].   These results likely indicate that movement occurs in discreet layers in the aquifer at different velocities with the upper layers thought to have the fastest transit times.

Jaclyn Gehring and colleagues at Union College recently used distinctive changes in the stable isotopes (oxygen) in river water to determine rates of groundwater recharge from the River to the Schenectady well field [7]. Her work shows that oxygen isotopes change over the course of a year and that these differences can be used to understand transit times of the River-derived groundwater.  They used measurements of Schenectady tap water to determine that water from the River travels at ~3.5 m/day and that water from the Mohawk reaches the Schenectady well field after 3.2 months (~96 days).  These measurements likely integrate a larger fraction of water in different parts of the aquifer compared to the radioactive isotopes and thermal plume.

So we know that water moves relatively quickly through the aquifer and there is a direct connection with the River.  As Paul Burwash reminded us in an accessible article on the Great Flats Aquifer in Clearwaters magazine:   “So it becomes necessary to understand fully how the aquifer functions in order to provide a means to detect pollution and develop a mitigation plan as the need arises. One thing is clear; what happens on the land’s surface is directly related to the Mohawk, and what happens in the Mohawk is directly related to the aquifer.”  [3]

2. The problem

I received a call the other day from Professor Carl George (Emeritus, Union College), who was the chair of the Glenville Well-field Protection Committee – a committee that spent several years thinking about the water supply for Glenville. I was also on this committee so I had some familiarity with the well field and its operation. When we were writing the report, we were concerned with threats to water supply that included flooding and potential surface spills from train derailments adjacent to the well field (a topic I recently covered in this post). We submitted our report to the Town in 2014, and in this report we suggested a number of strategies that the town of Glenville may follow to protect and optimize the well field [4]. Carl noted that there was a concern now (June 2020) because water levels in the wells were very low and pumping rates high.

With this in mind, I contacted Dave Ferris, operator of the Town of Glenville well field, and he confirmed that the water levels were probably about 5-7 feet lower than normal judging on the level in the observation well. He also noted that pumping is going at a brisk pace. In early June 2020 they pumped a record daily maximum of 4.9 million gallons a day (mgd), which is the highest they’ve ever produced, and well above their average annual daily mean of ~2 mgd [14]. He thought production was high because it has been very dry and people are watering their lawns, filling pools, and continueing with other warm-weather activities that demand water.  He also noted that some of the water from the Glenville well field goes to the towns of Ballston and Clifton Park. These water transfers are relatively new and they represent transfer of water through what is referred to as the interconnect system, which allow municipalities to share water when there are issues or problems.  In 2019 the Town of Ballston purchased ~85 million gallons from the town of Glenville [15].

Tom Coppola, commissioner of public works for the Town of Glenville, told me that the levels of water in the wells were as low as they have ever been since observations began back to the early 2000s.   His concern was that the operation of the locks on the Mohawk – or lack of operation - had affected groundwater levels. Specifically he noted that because Lock 8 was not in place, the pool of water in the Mohawk River that surrounds three sides of the Glenville well field was considerably lower (see map). Tom also noted that a number of other municipal water managers were very concerned about water levels elsewhere on the Great Flats Aquifer.  He said “We’ve never seen this” and noted that a request was made to the Canal Corporation to look into the possibility of lowering the removable dam at Lock 8 to produce the pool of water that would quickly aid aquifer recharge.  He was later told that under current conditions, this could not be done.  Note that this request was for the removable dams to be lowered to protect groundwater resources, not navigation.

Thus three issues appear to be intersecting in what may be the beginning of a crisis:  low rainfall, removable dams up and out of the water, and heavy water use typical in the summer. 

View west of the Mohawk River from the I-890 bridge on 18 June 2020. Photo: Thomas Copolla

3. The 2020 Water deficit and 2020 Pandemic intersect

It has been a dry spring and early summer in eastern NY and New England.  precipitation has been well below normal for the last 90 days, seven-day stream flow is much below normal (<10% for many stations in the Mohawk), and groundwater levels are also below normal [6]. There are mandatory water restrictions in several towns in MA (Ipswich, Georgetown) and dry conditions have resulted in an especially high number of wildfires in the state of Maine (500 in the spring of 2020) [6].

I recently wrote of the effect of Covid-19 on delayed opening of the Erie Canal, and it seems possible that parts of the Mohawk section will not open at all this year.  Canal issues are complicated, but suffice it to say that the Canal Corporation is not set up to just open a few locks: opening is an involved process where full access is needed with boats and coordination between locks and dams.  In a realistic scenario, the lower Mohawk may not be opened until mid-July at the earliest, so there is another month to worry and to watch and see if water levels in the aquifer continue to fall. Undoubtedly they will (see graph below).

One thing is clear: operations of the Canal directly affect groundwater levels in the Great Flats Aquifer. In a typical summer, Lock 8 and Lock 9 would be functional and the removable dams would be in place.  Lowering these dams raises water levels in the river by 14-15 ft, and this in turn raises adjacent groundwater levels.  A major issue for the Great Flats aquifer is that the removable dam at Lock 8 is not in place.

Operation of Lock 7 and the permanent Vischer Ferry dam also affect river levels in part because an extra 2 feet is added to the Schenectady pool when the flashboards are annually installed on the top of the dam.  This year everything is different due to the Pandemic and the pause in Canal operations. 

4. Contaminant concern

The combination of low flow conditions in the Mohawk and high rates of pumping may cause the wells to draw water from farther afield.

When the dams are in place and we have “full-pool” conditions, almost all the significant recharge in the aquifer needed for municipal wells is directly related to infiltration from the Mohawk River.  However, when the dams have been raised an important fraction of the water is drawn from aquifer “storage” – water that is only occasionally tapped.  When water is taken from storage water levels in the aquifer decline [2,8].  In this case, the cone of depression (area of drawdown of the top of the groundwater surface) for the Schenectady well field is thought to extend by as much as 1000’ upstream of Lock 8  [2, p. 71]. 

Studies of water temperature and well levels have shown that when the removable dam at Lock 8 is up, the Schenectady wells draw water from a layer that extends under the Mohawk to the north (Scotia) side [2,8].  This is important because it means that without the dam at Lock 8 and its associated pool of water, the Schenectady well field draws water from the aquifer on the other side of the Mohawk because transmissive gravels extend in a layer below the river. 

Headed for trouble.  Water level in the Schenectady monitoring well used by the USGS to monitor water levels in the Schenectady well field (well USGS 424859073585501 Local number, SN-730, Schenectady NY – see map for well location) [data 9].  The median line (black dotted) is the calculated median daily value for the 15 or 16 years on record.  The data for the 2020 water levels (blue solid line - through 17 June 2020), include data since late February that are provisional and subject to change. Note that annually the lowest median value is on 1 October.  What is important here is that in mid-June 2020, the level in the well was below the lowest annual median level. 

Some history of contaminated groundwater in the Great Flats Aquifer is important to consider.  Although the Great Flats Aquifer is the sole-source aquifer for Schenectady County, the area has supported considerable industrial activity and careless discharge of contaminants into the ground was not uncommon in the past.  One major suite of contaminants in groundwater in industrial areas is Volatile Organic Compounds (or VOCs) that include Tetrachloroethylene (PCE), Trichloroethylene (TCE), and Toluene – all used as solvents and for cleaning.  These compounds are known to cause cancer and they produce a number of other adverse health effects [10,11]. We need to do everything possible to keep these contaminants out of our well water, and considerable work has gone into doing just that.

In 2001 Alpha Geoscience conducted a hydrologic evaluation of groundwater use and contamination in advance of the construction application of the proposed Glenville Energy Park plant in Scotia (which was not built) that was to be a large consumer of water from the aquifer (Schenectady well field) [8].  There was concern that the additional draw of ~2 and 4 mgd from the Schenectady well field by the proposed plant may have adverse affects on quality and quantity of water from the aquifer. A particular concern was TCE.

In a review of water quality data, Alpha Geoscience provided a summary of TCE and Toluene contamination of the Schenectady well field.  Reported TCE levels between 0.4 and 1.0 ug/l (below the NYS drinking water standard of 5 ug/l) were reported in Schenectady water between 1990 and 1997 [8].  All four of the Rotterdam wells had measurable TCE in the analytical window between 1992 and 1999.  Remember that these are contaminants in wells on the south side of the Mohawk, and the nearest known TCE contamination is in areas in Scotia on the north side of the River.  A suspected source was somewhere in the Scotia Industrial Park.

Alpha Geoscience also reviewed data that showed pumping tests indicating that the cone of depression of the Schenectady well field extends north of the Mohawk River when the River makes limited contribution, which is the situation when the river stage and volume are affected by removal of the dam at Lock 8 and no flashboards on the Vischer Ferry dam at Lock 7 [8].  They concluded that the Schenectady wells draw from the aquifer north of the Mohawk during unusually dry climatic conditions and very high pumping rates – a situation we are in now.

The Defense National Stockpile Center Scotia Depot Site (or “Scotia Depot” site) in the Scotia Industrial Park is a NY State Superfund site due to TCE contamination [12,13]. In 2011 Schenectady County reached an agreement with the US General Services Administration to clean up a TCE plume from the Navy Depot[13]. The agreement was to address the plume at its source, but a considerable amount of the TCE had left the site and is traveling southeast to the Mohawk.  The GSA (Navy) spent $15 m remediating the plume at the source [5]. According to Schenectady Metroplex statement in 2011:

“The primary concern is a groundwater plume containing trichloroethylene (TCE) that has been identified at a depth of 70 feet below the surface of the Depot property. The plume, according to tests done by NYSDEC, is within a portion of the Great Flats Aquifer protection zone and is headed toward the Mohawk River. The aquifer is the multi-billion gallon underground water supply providing high quality drinking water to almost all Schenectady County residents.” [13]

The importance of the aquifer and the cleanup was noted in the NYS DEC Record of Decision in 2010: “Groundwater near the site is used as a potable water source by five Schenectady County municipalities (City of Schenectady, Village of Scotia, and Towns of Glenville, Niskayuna, and Rotterdam). Site related contamination is impacting a sole source aquifer that is used as a source of potable water. The remedy must address the impacts of the site to the groundwater aquifer.” [5]

The Alpha Geoscience report suggested that the Navy Depot TCE contaminant plume is to the west of a groundwater divide and therefore is not a threat to the Schenectady well field if pumping rates increase (as proposed in the Glenville Energy Park application).  However, they did not indicate the potential source of the TCE (and toluene) that ended up in the Schenectady and Rotterdam well field in the 1990s.

They also suggest that the groundwater divide is in part caused by the effects of Lock 8 on groundwater levels, and that groundwater flow direction is “strongly influenced by the change in river elevation at Lock 8.”  What happens when Lock 8 in not in place? Has part of the plume migrated southward?  Or is there another TCE source in the Scotia Industrial Park that has managed to get into the groundwater and travel across the river? 

5. Where we stand – action needed

It may very well be that we are at a pivotal point with respect to groundwater use and drawdown of the aquifer. The Great Flats Aquifer has enormous capacity, but it is very important that we consider how the rate of pumping affects the level of wells and how the expanded cone of depression might begin to draw contaminated water.

This current situation should be a concern for Schenectady and Rotterdam because they have detected TCE in well water in the recent past [8].  It should also be a concern for Glenville because if those wells draw water from father afield, they may intersect impaired groundwater from the Schenectady International site – also contaminated by VOCs [16,17] - to the northwest (see map).  This possible contamination route was evaluated by the Glenville Well Field Protection Committee back in 2013 [4], but those discussions considered full-pool summer conditions, not no-pool summer conditions with high average daily withdrawals.

A simple solution is for the Canal Corporation to consider the act of lowering the dam at Lock 8 to be “urgent and essential.”   Establishing the Lock 8 pool would quickly affect groundwater levels in the Schenectady, Rotterdam, Scotia, and Glenville well fields.  The Canal Corporation would need to somehow work out a mechanism to lower the removable dam (and add flashboards on the Vischer Ferry dam) while the Pandemic and work restrictions continue.  Time is not on our side: this situation needs to be addressed now because in August or September it may be too late to attempt an intervention.

---

This and other Notes from a Watershed are available at: https://mohawk.substack.com/

---

Further reading

[1] Waller, R.M. and Finch, A.J., 1982. An atlas of eleven selected aquifers in New York (No. 82-553). US Geological Survey,

[2] Winslow, J.D., Stewart, H.G., Johnston, R.H. and Grain, L.J., 1965. Ground-water resources of eastern Schenectady County, New York. NY Water Resources Commission Bull, 57, 148 p. (Link Here).

[3] Buzash, Paul R. 1999.  Understanding and protecting the Great Flats Aquifer and the Mohawk River, Clearwaters, v. 29, n. 4, p. 14-17.

[4] George, C., and the Glenville Well Field Protection Committee.  2013.  Advisory report on Protection of the Glenville Well-field.  Unpublished report available from the Town of Glenville. (Link Here).

[5] NYSDEC Record of Decision (March 2010) - Defense National Stockpile Center Scotia Depot Site, State Superfund Project Town of Glenville, Schenectady County, New York Site Number 447023 (Link Here).

[6] Northeast Drought Update - Posted June 11, 2020 - Northeast Regional Climate Center at Cornell (Link Here).

[7] Gehring, J., Stahl, M., Gillikin, D., Verhyden-Gillikin, A., 2020.  Streamflow capture along the Mohawk River: Determining transit time to municipal well-field.  In Garver, J.I., Smith, J.A., and Rodak, C. 2020. Proceedings of the 2020 Mohawk Watershed Symposium, Union College, Schenectady, NY, March 20, 2020, v. 12, p. 24-27. (Link to abstract volume here).

[8] Alpha Geoscience report from 2001 on the TCE plume in the GFA.  Alpha Geoscience, 2001, Hydrology of the Great Flats Aquifer in the vicinity of the Glenville Energy Park site.  Confidential document prepared for Earth Tech Inc, Albany NY (appendix to NYSDEC documents) (Link Here).

[9] Data for groundwater levels of the Schenectady wellfield monitored by the USGS (graph) (Link Here).

[10] VOCs in groundwater, general overview by Bruce Lindsey at the USGS (Link Here).

[11] Trichloroethene (TCE) at the National Cancer Institute (Link Here).

[12] Trichloroethene remediation at Defense National Stockpile Center Scotia Depot Site, NY State Superfund Project Published Aug. 1, 2019 [Link Here].

[13] Trichloroethene to be addressed by General Services Administration.  Schenectady Metroplex.  Published 12 October 2011. (Link Here).

[14] The town of Glenville annual Water report (2019) (Link Here).

[15] The town of Ballston annual water report (2019) (Link Here).

[16] EPA site entitled: “Hazardous Waste Cleanup: SI Group Incorporated in Rotterdam Junction, New York (US EPA).”  Site was remediated by Consent orders in 1993 and 1997.  New York State 373 Permit was issued March 27, 1998 and renewed in 2003. The EPA notes that the site is contaminated with Volatile Organic Compounds (VOCs) (benzene, ethylbenzene, xylene and toluene); and Semi-volatile Organic Compounds (SVOCs) including Polycyclic Aromatic Hydrocarbon (PAH) and phenols.  This site is directly across the Mohawk (to the northwest) of the Glenville Wellfield. (Link Here).

[17] NYS DEC Fact Sheet for Schenectady International - Rotterdam Junction Operable Unit 1 - SITE No. 447001.  October 2018. (Link Here).