1. Impacts of Whole House Reverse Osmosis Systems on Septic Systems
Presented by:
Kevin Koepenick, L.E.H.S.
Baltimore County Department of Environmental Protection and Sustainability

2. Overview
Natural Ground Water Quality and Typical Water Treatment in Piedmont
Observed Trends in Wells Contaminated by Road Salt
Options for Addressing Road Salt Comtamination
Addressing Wastewater from R/O Systems
Today I am going to be discussing the problems I am experiencing with a growing number of homeowners whose wells have been contaminated by road salt and they have elected to install a treatment system to address the problem. What the homeowner is usually not aware of is that the typical water treatment technology for removing the salt, reverse osmosis, generates a large volume of wastewater that must be addressed.
But before I get to that issue, I wanted to set the stage of where we are seeing this problem in Maryland and how the natural geology is affecting this issue. I will also present the trend in complaints that we have received in Baltimore County detail you the problems homeowners are experiencing this this type of contamination.
Then we will discuss the reverse osmosis (R/O) systems that many homeowners are installing and the impacts they can have on an onsite sewage disposal system.

3. Maryland is underlain by a wide range of geologic formations spanning 5 different Physiographic Provinces. The focus of this talk with be on the areas underlain by the “hard rock” aquifers west of the sedimentary formations of the Coastal Plain. As we will discuss, wells in these crystalline aquifers are particularly susceptible to contamination relative to Coastal Plain aquifers since they are unconfined, relatively shallow, and are directly impacted by land use in the vicinity of the wellhead.

4. Occurrence of Ground Water in the Piedmont
Water Table
Well #2
Saprolite
Well #1
Stream
Stream
Fracture
This slide depicts a typical cross-section of a Piedmont Aquifer. There is usually a relatively thin layer of soil (ranging from 4-8 feet) that is underlain by weathered bedrock known as “saprolite” and beneath that is the consolidated bedrock. The top of bedrock is usually no more than feet deep. Although most bedrock wells are completed into the bedrock at depths between 200 and 500 feet, there is actually very little water storage in the bedrock fractures themselves. Most of the water is stored in the saprolite which sits on top of the bedrock. So, the water from the wells is transmitted through the fractures from the saprolite. This water is recharged seasonally by precipitation—most of which occurs in the cooler months (Nov – March). The net effect is that contamination that makes it through the relatively thin layer of soil can have a direct impact on the water coming from a given well.
Bedrock

5. Roughly 10 % or 85,000 Baltimore County Residents
Rely on an Estimated 36,000 Private
Well Systems
In Baltimore County, there are about 36,000 homes and businesses served by well and septic systems. The tan areas indicate suburban and rural areas served by well and septic systems. The light blue areas are served by metro water and sewer. The majority of the rural areas are underlain by the bedrock aquifers of the Piedmont.

6. Natural Water Quality in the Piedmont
Concern
Treatment Option
Low pH
Neutralizer (Calcium Carbonate)
Iron
Softener (Ion Exchange)
Manganese
Radium
In 1998, the Maryland Geological Survey completed a detailed study of water quality in Baltimore County. Overall, water quality from wells sampled throughout the county was found to be of generally good quality and safe for human consumption. However, due to the common occurrence of elevated iron, manganese and the generally low pH ( ) of the groundwater, many water supplies have treatment systems which consist of a neutralizer and water softener. There are also areas where naturally occurring radium is also present and levels of concern. Radium is also effectively removed by water softening systems. The result is that many homes are equipped with these systems whether they need them or not.

7. Historically, our office occasionally receives calls from residents with water quality concerns and by and large the typical problems were associated with leaking underground fuel tanks. However, in recent years we have seen a tremendous increase in the number of documented chlorides cases over the last decade. This data shows 145 cases in just the last seven years—78 in just the last two years alone. These are typically complaints from property owners with private well water supplies that have experienced serious issues with the usability of their water.

8. The red triangles on this map show the approximate location of each of these chloride cases. They are fairly well distributed but not surprisingly concentrated along major roadways.

9. Domestic Well Complaint Concentrations 2014-2015
Chloride Range: 341 – 1361 ppm Chloride Ave: 567 ppm
Sodium Range: 45 – 575 ppm
Sodium Ave: 196 ppm
Sodium/Chloride Ratio: 9 – 60%
The range of chloride levels from these complaints in just the last year is from 341 to over 1300 ppm with an average of 567 ppm
Sodium ranged from 45 to nearly 600 pp with an average of almost 200 ppm.
I was surprised by the sodium levels since I had assumed that most of the it would be bound up in the soil but you can see by this data that some of the data showed the sodium to chloride ratio at up to 60%. So perhaps there is a threshold of sodium that the soil can hold and the excess is moving along with the chlorides.

10. Typical Homeowner Complaints
Replacing Water Heater Every 2 years
Frequent Leaks (from corrosion)
Can’t Drink Water Due To Taste
Dry, Itchy Skin / Hair Loss
I have heard story after story of homeowner hardships due to the constant plumbing problems they have faced. The most common complaint is the on-going need to replace their water heaters, frequent leaks (especially in copper plumbed houses) and poor taste of their water. I have also heard about dishwashers etching their glasses, and rust staining on their clothes from the washing machine, as well as dry itchy skin and hairloss.

11. Corrosion
Here are a few examples of what the elevated chloride levels have done to people’s plumbing and fixtures. Note the tell tale aqua-blue of copper oxide that forms from the corrosion of the copper pipes.

12. What is the Cause?
Rock salt is not naturally occurring in the region so the only source of the salt has to be man-made. And the most likely sources are the application of road salt, and discharge from water softeners and septic systems

13. Increased road salt use & Cl – in streams
Rural rivers feeding Baltimore water supply reservoirs
3x increase [Cl–] since 1960
This graph shows the of total US highway salt sales overlain with a graph showing the increase of chloride concentrations in Baltimore Metro water supply through time. The increase in salt usage correlates well with the change in the chloride concentrations
Jackson R. B., Jobbágy E. G. (2005) PNAS; Kaushal & Belt (2012) Urban Ecosys

14. Road Salt Usage in Baltimore County
Fiscal Year
Storm Events*
Est. Lane Miles Maintained by SHA
Tons of Salt Applied by SHA
Tons of Salt Applied by SHA/Lane Mile/Event
Est. Lane Miles Maintained by BC**
Tons of Salt Applied by BC
Tons of Salt Applied by BC/ Lane Mile/Event
2000 6
1558
31931
3.4
6400
41668
1.1
2001
26741
2.9
6430
47051
1.2
2002 2
1561
14128
4.5
6465
24105
1.9
2003 15
59054
2.5
6517
102042
1.0
2004 10
47420
3.0
6537
73380
2005
40804
2.6
6567
72232
2006 4
22878
3.7
6587
33947
1.3
2007 7
37697
6615
62858
1.4
2008
30901
2.8
6640
47806
2009
1577
37210
2.4
6663
43632
0.7
2010
55027
3.5
6682
94477
2011 12
1674
29592
1.5
6694
82504
2012
11570
6701
10611
0.3
2013 18
25666
6711
36902
2014 20
97094
6722
125309
0.9
2015
49929
1.7
AVE
10.06
38603 3
* According to SHA data
** Lane Miles Estimated by Multiplying Linear Road Miles by 2.5
In this table I summarized the salt usage in Baltimore County by both SHA and DPW, and calculated the amount of salt applied per lane mile per snow event. You can see that the amount of salt usage varies considerably year and the total lane miles are increasing by about 5-7% during the 15 year period. Overall, there were about 10 snow event/ year and the combined salt application averaged almost 100 K tons per year

15. Typical Softeners
1 to 3 regeneration cycles/ gallons / cycle over 1-2 hours
Use about 1 lb salt for gallons of water usage
Studies show softeners will increase chloride levels in septic tank from mg/l to mg/l
Compare this to a typical water softener usage with has 1-3 regeneration cycles per week discharging about 10 lbs of salt a week. Some studies have shown that these softeners increase the chloride levels in septic tank from around 70 mg/l to several thousand mg/l and at any given time could be as much as 10,000 mg/l.

16. Water Softeners Assuming 18,000 Systems in BC Using 50 lbs Salt/Month
= 5,400 Tons of Salt/Year
= 5% of Total Road Salt Applied
For a rough idea about how much salt might be coming from individual softening systems that have a brine discharge, I assumed that 50% of the homes has a softener and used about 50 lbs of salt a month. This equate with only about 5% of the total salt that is being applied to the roads each year.

17. Septic Systems Assuming 36,000 Systems in BC
Using mg/l Cl-
= 80 Tons of Salt/Year
= < 0.1% of Total Road Salt Applied
For a typical septic system without a softener, this number comes out to less than 1 tenth of 1 percent of the total amount applied to road salt.

18. Common Beliefs
Salt discharge from softeners will hydraulically overload the drainfield
Salt will reduce the permeability of the drainfield
Salt will kill off the “good” bacteria in the septic tank
The impacts of brine discharge to septic tanks and drainfields has been heavily debated in recent years. There are studies that show that well maintained softening systems have relatively low impact on the hydraulic load and minimal concerns with regard drainfield permeability and biological activity in the septic tanks. However, there also many reports to problems with septic systems receiving backwash from softening systems. It isn’t clear as to whether these problems are from malfunctioning softening systems or unusual cases.

19. Common Beliefs
Salt reduces settleability in the tank thereby increasing solids moving to the drainfield
Salt will reduce the effectiveness of an ATU and void manufacturer warranty
Salt will increase corrosivity of the wastewater affecting the life of the concrete components
It seems logical that the introduction of large volumes of concentrated salt water would reduce the settlement of solids in the septic tank, due to the heavier salt water sinking and stratifying the tank but again, it may depend on the volume and frequency of backwash cycles as to how much of a problem this really is.
Certainly, manufacturers have made it clear that they do not want the brine backwash being introduced prior to their pre-treatment units.

20. How to Manage Address Wells Impacted with Chlorides
Remove From Water Using Reverse Osmosis (Very Inefficient)
Drill New Well
For those whose well have already been contaminated they basically have two options: 1) install a reverse osmosis system and treat the water or 2) drill a new well—this assumes there is enough land area to accommodate and you also have the possibility of drilling a dry hole or one with considerably less well yield (i.e., no guarantee)

21. Problems Related to R/O Treatment
High Cost ($15 K - $20 K)
Very Inefficient (1-3 gallons of wastewater for each gallon of clean water)
Hydraulic Capacity Concern for Septic Systems
Soil Permeability Concerns
For those residents who choose to treat their water, they are faced with not only the relatively high cost of the treatment system itself but ongoing maintenance costs ($1000 / year); potential loss in water pressure (need storage tanks and booster pumps) and the extra water usage may overload their septic resulting in a need to expand the system.
As I have mentioned all the suppliers of whole house R/O systems in my area are telling me that the best efficiency you can reasonably expect from these system is 1:1 (one gallon of wastewater for every 1 gallon of “salt free” water)
So where does all this wastewater go? That is what we have been grappling with for the last few years.

22. Typical OSDS Design 150 gallon/day/bedroom
Loading Rates of gal/ft2/day
Assumes Septic Tank Effluent
Typical 4 Bdrm Home = 1500 gal Septic Tank and (2) 75’ trenches 2’ wide and 4’ deep
A typical conventional residential design in Maryland assumes a peak daily load based on 150 gallon/day/bedroom and a loading rate between 0.6 and 1.2 gal/ft2/day. So for a typical 4 bedroom home this might equate to a deep trench design consisting of a 1500 gallon septic tank and (2) 75 ft trenches 2 ft. wide and 4 ft deep assuming 2 ft of sidewall absorption.
So if we take the same house and we put on a whole house RO system---do we need to double the size of the septic system?

23. Questions Raised by R/O Usage
Should discharge go to Septic System?
If Yes, How to size field system?
What are the Design Criteria?
Should the R/O wastewater be separated?
Should the discharge go to the the septic system? If so, how should we size the system Do we need a septic tank ? Can we use a different loading rate? ? Should we separate the wastewater ?

24. Maryland Regulation
Discharge to surface must be individually permitted by MDE
Discharge to Septic allowed provided adequate capacity
Bypass Tanks Recommended
1/3 increase in L.R. Recommended
In 2015, MDE provided official guidance that stipulated that backwash from treatment units may only be discharged to the ground through an individual permit issued by MDE or through an adequately sized septic system. The guidance recommended that the brine waste be discharge directly to the drainfield (around the septic tank) or discharged to a separate drainfield altogether
This memo did not provide any guidance about the how to determine if a system was adequately sized.
In 2017, MDE provided draft guidance with further details on this issue and recommended that where trenches are to be used for wastewater exclusively from water treatment systems that the loading rates could be increased by 1/3 (basically following the Tyler Chart for loading rates using pre-treated effluent).

25. Other States Rhode Island allows direct discharge to surface or OSDS
Massachusetts and Connecticut prohibit discharge to OSDS
Delaware – Need waiver to send to OSDS or can send to French drain
Penn. – Allows discharge to OSDS
Virginia – Allows discharge to OSDS
Rhode Island allows direct discharge to the surface but recommends it be to a dedicated drywell whenever possible an provides guidance on setback distances wells and adjacent properties
Massachusetts and Connecticut prohibit direct discharge the septic system. Delaware requires a waiver to send to the septic but will allow discharge to got to a French drain without a permit
Pennsylvania and Virginia are similar to Maryland with no specific guidelines that address this issue.

26. Study by Siegrist (1987) LTAR
Let’s look at what the research tells us. One of the fundamental principles for determining soil loading rates for OSDS is the concept that increasing levels of BOD and TSS reduces infiltrations rates in soils over time. This graph is from a study by Robert Siegrist and it shows the results of an experiment where 4 different strengths of wastewater were applied over the same soil A – High Strength Wastewater; B – Typical Domestic Septic tank Effluent; C- Graywater Strength Effluent and D—tap water. You can see from the graph that over a period of months to years, the initial infiltration rates drops off considerably until they reach was is referred to as a long-term acceptance rate. But notice that the line for “D” wastewater (tap water) does not change because the hydraulic permeability is not affected by the “clean” wastewater.
Tyler and Converse took this a step further and recommended specific loading rates based upon actual soil type, and structure. These loading rates assume that the wastewater being applied is either typical “septic tank effluent” (domestic strength effluent) or pre-treated effluent. In their 1989 paper, Tyler and Converse agreed with Siegrist’s findings and concluded that highly pre-treated effluent could applied at higher loading rates—possibly as high as the saturated hydraulic conductivity of the soil (Ksat)

27. Siegrist Article (2006) from Small Flows Quarterly
In 2006, Siegrist published an article in the Small Flows Quarterly challenging the industry to move from the prescriptive code requirements based on a standard “perc” tests to a loading rates based on soil classifications and Ksat measurements. He argues that the standard perc test is a “crude” method to base design on and suggests that maximum loading rate be 5 to 10 % of the soils Ksat value as measured from a more precise method. The chart above from his paper shows 3 different soil classifications with Max Daily Loading Rates ranging from 12.5 to 0.25 gal/ft2/day depending on the soil type.
The USEPA Onsite Wastewater Treatment Manual also references this finding from Siegrist and others stating that wastewater with low BOD concentrations can be applied as soil at rates from 2 to 16 times typical hydraulic loading rates for domestic septic tank effluent.

28. Effect of Sodium on the Soil
Many studies show decrease in H.C. for clay soils
Tyler et al.(1978) say softeners should not be a problem for H.C.
One study suggests using K Cl vs Na Cl
May not be an issue for deep trenches in saprolite
So what is the impact of sending brine laden water to the soil absorption system? The chloride ion does not react with anything and moves readily to the groundwater. The sodium ion, however, has been shown to cause reduction in Hydraulic Conductivity particularly for clay-rich soils due to soil dispersion. This is contrary to an often cited paper by Tyler et. al. (1978) that concluded that brine from softeners should not have an adverse effect on soil permeability assuming that there is sufficient amount of calcium and magnesium in the effluent to prevent dispersion from occurring. One researcher suggested that using K Cl in softening systems may significantly minimize the risk of soil dispersion but K CL is generally a more expensive product than Na CL and is therefore not typically used in road salting operations.
I think the risk of significant problems related to reduced H.C. should be considered with increasing clay content of the soils. I don’t expect to see many problems in deep trench systems that we design in the saprolite where perc rates are generally less han 15 min/in.

29. Current Approach in B.C.
Evaluate the existing system, flows, and capacity
Recommend Water Metering
Recommend Separate Trench for R/O Wastewater with Obs. Ports
Design on 2 x L.R. for STE
At this point in Baltimore County we are still handling these issues on a case-by-case basis. The first thing we do is check our records on the age and design of the system and if necessary have a sewage disposal contractor conduct a detailed inspection of the system to determine if functionality and remaining capacity. I generally interview each homeowner and discuss there water usage and the additional volume of wastewater than can be expected if they install a whole house R/O.
If necessary we will conduct percolation tests and determine if they have enough room for a separate disposal system to handle the R/O wastewater. In many cases, the property owners do not want to install a separate trench system due to the extra costs. In those cases where they want to send the wastewater to their septic, I strongly urge bypassing the septic tank and water metering and monitoring of the water usage and sewage levels in the trenches or drywells. If they begin to see hydraulic problems we will work with them for alternative disposal options.
Where they install separate trenches we design bases on 2 times the loading rate that we would typically use for a conventional septic system. This results in halving the system size needed for domestic seweage (cheaper costs). From the research I reviewed, we may be able to increase the loading rate 5 to 10 times greater than we would for a typical septic system, with more conservative rates for soils with increased clay content.
At this point I see further research and discussion is needed but consideration to be given to changing the regulations in Maryland to either prohibit wastewater discharge to septic systems from water treatment systems or to provide better design parameters for addressing this growing issue.

30. Questions ?