Upload
others
View
6
Download
0
Embed Size (px)
Citation preview
SEPTIC TANK WASTE STRENGTH AND SAMPLING ONSITE SYSTEMS:
THE NUTS AND BOLTS
Barbara J. Rich, Dan Haldeman, Todd Cleveland, and Jill Johnson1
This paper reviews the wastewater characteristics of twenty households and compares the effluent
quality from the treatment unit discharge pipe versus the pump chamber following the treatment
unit. This paper also illustrates the procedures and the reasons for their use as defined by the La
Pine National Decentralized Wastewater Treatment and Disposal Project (La Pine Project).
The existing studies or protocols developed for sampling onsite wastewater systems was scarce
when the La Pine Project began work. The primary source of information was the procedures used
in sampling municipal wastewater treatment plants and industrial pretreatment processes. Onsite
wastewater treatment systems turned out to be enough different from these situations that the La
Pine Project Team defined some alternative methods of obtaining representative data and also
provided initial testing of a hypothesis that grab sampling produces nearly identical results for most
parameters as composite sampling.
Sample parameters
Defining the sample parameters depends largely on the local situation and contaminants of concern.
The La Pine Project study is located in an area where nitrogen contamination is a concern due to the
presence of a sole source drinking water aquifer, the shallow depth to the water table, and rapidly
draining soils. Therefore, one of the most important sample sets to take is the nitrogen species. This
includes total Kjeldahl nitrogen, ammonium, and nitrate-nitrite at a minimum. A separate analysis
for nitrite may be warranted if there are difficulties troubleshooting the advanced treatment process.
The separate nitrogen species show how well the advanced treatment system can accomplish the
different stages of the nitrification/denitrification process.
The 5-day bio-chemical oxygen demand (BOD5), total suspended solids (TSS) and the bacteria
analyses provide a basic characterization of wastewater quality. The chloride analysis provides a
means by which to account for dilution by precipitation or irrigation in systems that are open to the
atmosphere. Chloride data can also provide the first indication of the impacts of residential sewage
dispersal in monitoring wells. Often the chloride will show up in detectable quantities before the
nitrogen is apparent. Total alkalinity is a good diagnostic parameter because the nitrification process
for a gram of ammonia consumes a maximum of 7.14 mg. Typically, less alkalinity than the
theoretical quantity is used because of the nitrogen used in building cellular structures, but the 7.14
mg/L provides a good basic rule by which to determine if the nitrification process is alkalinity
limited. (Crites and Tchobanoglous, 1998; Burks and Minnis, 1994)
1 Barbara J. Rich, La Pine Project Coordinator, Oregon Department of Environmental Quality, c/o Deschutes County
CDD, 117 NW Lafayette Ave., Bend, OR 97701, 541-617-4713, [email protected]; Dan Haldeman,
Supervising Sanitarian, Todd Cleveland, Registered Sanitarian, Jill Johnson, Sanitarian Trainee, Deschutes County
Environmental Health, 117 NW Lafayette Ave., Bend, OR 97701, 541-388-6575
Fats, oils and grease is taken from the septic tank but not any other location in the treatment stream
because the project team uses this parameter primarily in the evaluation of the septic tank effluent
against the definition of residential waste strength currently in the Oregon rule (Oregon DEQ, 2000).
This parameter is also used when troubleshooting systems’ performance, however, the advanced
treatment systems were not required to reduce fats, oils and grease as part of the demonstration
project. Table 1 lists the parameters taken during wastewater sampling for the La Pine Project.
Table 1. List of parameters for the La Pine Project.
Parameter Septic tank effluent (STE)
All other
wastewater sites
BOD5
Total Alkalinity
TSS
Total Phosphorus
TKN
Ammonia as N
Nitrate+Nitrite as N
Chloride
Fats, Oil & Grease
E. Coli
Fecal Coliform
Preparation for sampling
The preparation for sampling is as important as the sampling activity itself. We have found that the
more details that are taken care of during the preparation period, the fewer mistakes that are made
during the sampling event. Obviously, sampling is expensive so fewer mistakes equates with greater
integrity of the dataset because of the increased confidence level in the quality of the work
performed and the fewer missed sample events.
We set up a small lab in one of the Deschutes County buildings as a place out of which to base the
sampling activities. This space was separated from the Environmental Health Division offices in
order to provide storage for sampling equipment and supplies, an area for equipment maintenance,
and adequate ventilation to minimize odor complaints from the sample filtering process and clean up
operations. We also tried to complete at least initial clean up in the field to avoid bringing dirty
equipment into the building.
Long-term sampling warrants using portable meters instead of test kits when taking field parameters
to ensure consistency and lower long term operating costs. We calibrated the pH and
DO/conductivity meters at the beginning of each sampling day. This schedule is more stringent than
required by the equipment manufacturers, however, we felt this approach was more defensible over
the long term, especially as the sampling events became less frequent towards the end of the project.
The DO/conductivity meters are also calibrated to the altitude of the sampling site upon arrival
because the elevation of the lab (located in Bend) is significantly lower than the south Deschutes
County area.
We separate the sample bottles for each sample site into sets and record bottle numbers ahead of
time on the field/chain of custody sheets. The bottles are bound into sets using rubber bands or a
dedicated bucket for each location. Creating bottle sets before the start of the sample day ensures
that there are enough bottles to complete the day’s sampling and recording the bottle numbers before
arrival on the site dramatically reduced the number of transcription errors. Again, fewer errors
translates into less wasted resources.
Preparation for sampling on the site
Two people typically comprised the wastewater sampling team with each person performing the
same task (reading meters, taking samples, opening components, etc.) to ensure consistency in
procedure. This division of duties was also the most efficient in terms of the amount of time spent
on a site. For instance, the team member that usually operates the meters will set up the sample
table, the meters, set the bottles next to each sample location and takes field observation.
Meanwhile, the other team member is opening the component lids, taking out the sampling devices
and sets out the clean up station.
The sampling team used a small folding table on the site for the meters and sample preservation.
The table required frequent cleaning and sanitizing but the ergonomic advantages outweighed this
relatively minor factor.
Sampling procedures and equipment
Procedures – the sampling procedures are provided in the Quality Assurance Project Plan for the
project. The page outlining the wastewater sampling procedure is provided in Figure 1. Comparable
procedures are available for groundwater sampling, meter calibration, and lab follow-up after
sampling. We designed each procedure to be easily replicated and laminated for field use in the
event that additional personnel needed to be trained.
Our primary approach in sampling a wastewater treatment system is to sample from “clean” to
“dirty.” This approach means that we begin sampling at the point where the effluent passes beyond
our ability to collect a sample and work our way up stream. For example, if we are sampling a
packed bed filter system that includes a lysimeter in the dispersal field, we begin with the lysimeter,
then sample the effluent pipe of the packed bed filter and last, sample the septic tank. This ensures
that we do not inadvertently contaminate a downstream or “cleaner” sample with upstream or “dirty”
effluent.
Figure 1: Wastewater Sampling Procedure
Equipment
Apart from the fundamental question of where and how to sample, there is the even more basic
question of what kind of equipment should, or shouldn’t be used. Some of these conclusions were
drawn through trial and error and some were arrived upon as a matter of practicality or safety. The
sample collection tools used at different types of sample sites are summarized in Table 2.
Table 2: Sample collection device used to sample different effluent discharge points.
Equipment Sampling site Comment
Dedicated
Disposable
Teflon Bailer
Septic tank
effluent – gravity
discharge
Each site requiring a bailer will have a dedicated bailer that is
cleaned/sterilized after each use and replaced quarterly.
Dipper Septic tank
effluent –
screened pump
vault
Particular care must be taken to avoid disturbing the attached
growth on the side of the screened vault or on the floats
Treatment unit
discharge pipe
The dipper is used when the system discharges frequently
enough and in great enough volume to collect during a sample
visit. Basic protocol is not to force the system to discharge
prematurely.
Pump chamber
following the
treatment unit
This is the sample site of last resort because the bacteria data
from this site is not representative of what the treatment unit
is discharging.
Bucket Treatment unit
discharge pipe
The bucket is left at those treatment unit effluent pipes that do
not discharge frequently enough or do not discharge a large
enough quantity during a sampling visit. The bucket is left so
that it performs as a flow through cell until the sample is
collected. The collection period is typically 2-24 hours.
Lysimeter or
Sand Filter
Effluent
The bucket is left for a period of a day to several days and
used as a flow through cell. Typically lysimeter or sand filter
effluent is not discharged fast enough for the sample to be
collected during the sample visit.
All equipment is scrubbed and washed with detergent and then sanitized using a bleach solution.
The project team discussed the potential for the chlorine bleach to confound the sample results, in
particular the bacteria findings. However, the team decided that, while the risk was minimal, rinsing
the sampling equipment in the effluent before sampling and to always take the bacteria samples last
could further minimize it.
After sampling is completed, all sample collection equipment is cleaned and sanitized before leaving
the site. (The groundwater sampling equipment requires different clean up procedures that will not
be addressed here.) Three five gallon covered buckets are carried in the sampling vehicle for
detergent wash, rinse, and sanitizer. The dipper, bailer, bucket, and sludge judge are scrubbed,
rinsed, and sanitized before leaving the site. This ensures that the equipment meets the minimum
contact time with the sanitizer and is ready for use at the next site.
Upon return to the lab, the chloride samples require filtration prior to shipping for analysis. Normal
procedure is to filter in the field; however, filtration in the field requires the use of a peristaltic pump
and an in-line filter. This approach not only increases the quantity and kinds of equipment required
for sampling but also poses a significantly higher risk of the sampling personnel coming into contact
with wastewater. Septic tank effluent, in particular, and some treated effluents clog the in-line filters
quickly which causes significant pressure build up in the pump line. The result is often a high-
pressure stream of effluent and an unhappy sampling team when the hose clamps fail. As an
alternative, the sampling team uses a filter holder with a receiver and funnel for filtering in the Bend
lab. This filter is set up with a peristaltic pump providing a vacuum so that the effluent is pulled
through the filter rather than pushed. The main complaints with this method have been odor
complaints from neighbors in the building where the lab is located, which were significantly reduced
when the ventilation was improved. The advantages to this approach include a safer process and
lessened costs for materials.
Because most of the equipment is cleaned in the field, clean up in the lab consists primarily of
servicing any meters or other equipment that may need it, packing samples for shipping to the
analytical lab, completing paperwork and restocking supplies.
Personal Protective Equipment
The primary method of ensuring personal protection is care during sampling. Following that, gloves
and proper hand washing technique and eye/face protection are used to protect sampling team
members. Microbiologically, the foremost concern is the potential for effluent splashing or spilling
during sampling. We have found that using a dipper with a nine-foot handle has greatly minimized
the potential for sampling personnel to come in contact with effluent at the sites where dippers can
be used. Bailers can be slightly more risky to use, particularly as they are used almost exclusively
for septic tank sampling, but careful technique minimizes the potential for splashing. Certain
personnel prefer using a face shield and apron during sampling for greater protection.
Boots are particularly important if concrete riser lids are present or if the native wildlife warrants it.
During winter sampling, boots with insulated soles were essential protection from snow or ice.
Appropriate cold weather gear is essential because the wastewater sampling does not stop for
inclement weather, in fact, this data can be important in determining whether or not the treatment
units are sensitive to environmental conditions
Sample locations
Identifying the representative sample locations included a discussion as to the merits of different
approaches. The project team agreed to adhere to the primary goal of the La Pine Project (evaluate
the performance of advanced treatment systems for nitrogen reduction) when designing the sampling
approach. In this light, the septic tank evaluation, while a valuable study in and of itself, remained a
supporting effort to the primary goal. Here, the project team decided to focus on septic tank effluent
in order to best characterize the influent quality that the treatment units were receiving.
In this analysis, we compared the septic tank performance using total nitrogen instead of total
Kjeldahl nitrogen (TKN). The nitrogen in the septic tank effluent is comprised primarily of TKN
with minimal nitrate present. The project team wanted to look at the total amount of nitrogen
leaving the septic tanks to determine if that would be a better representation of effluent quality than
TKN. The statistics for TKN are indistinguishable from the statistics for TN and so the project team
decided to report the septic tank effluent quality using the TN results in keeping with the
performance standards for the innovative treatment system study. Table 3 provides a summary of
the septic tank effluent data obtained to date.
Table 3. Overall septic tank effluent statistics for 19 non-recirculating septic tanks in the La Pine
Project.
Parameter Mean Median
Standard
Deviation Minimum Maximum Count
BOD-5 (mg/l) 264 250 138 22 1000 379
Total Suspended Solids 85 61 106 0.5 1600 378
Total nitrogen 64.9 63.0 19.4 8.6 150 377
Oil & grease 35 28 28 2.5 280 367
Fecal coliform 1.676 x 107 1.700 x 10
5 7.757 x 10
7 2 7.700 x 10
8 379
log Fecal coliform 5.42 5.23 1.37 0.30 8.89 379
E. Coli 1.036 x 107 1.300 x 10
5 5.528 x 10
7 2 7.400 x 10
8 379
log E. Coli 5.26 5.11 1.35 0.30 8.87 379
Units for BOD-5, TSS, TN, and O&G are mg/L. Bacteria counts are in CFU/100ml.
The data over all the non-recirculating septic tanks in the La Pine Project suggests that Oregon’s
residential waste strength definition may no longer be valid. Oregon’s current definition states the
following:
"Residential Strength Wastewater" means the primary sewage effluent from a septic tank
which does not typically exceed the following parameters: Five-Day Biochemical Oxygen
Demand (BOD5) of 300 mg/L; Total Suspended Solids (TSS) of 150 mg/L; Total Kjeldahl
Nitrogen (TKN) of 150 mg/L; and Oil & Grease of 25 mg/L. Other contaminants may also be
present in the wastewater, however, they shall not exceed the concentrations or quantities
normally found in residential sewage. Effluent parameters are to be measured using
approved Standard Method or EPA procedures. (Oregon DEQ, 2000)
The nineteen septic tanks in the La Pine Project represent a diverse group of residents from single
young or retired persons to families of six. The sites include residents on long-term antibiotics or
chemotherapy drugs to people who live an organic lifestyle. While this diversity creates a difficulty
when trying to determine how the one-compartment tank population compares to the two-
compartment tank, the La Pine Project Team believes that the diversity is representative of the
population in general. Given that, Table 4 shows how the sample population compares to the
residential waste strength definition. The 9 single compartment tanks were sampled between 19 and
31 times each, with the average number of samples being 25. The number of samples taken from the
10 two-compartment tanks ranged between 13 and 19 each with the average number of samples
taken overall being 16. The difference in sample numbers between the two populations is primarily
due to differences in installation dates. The majority of the two compartment tanks were installed
later than the single compartment tanks.
Table 4. Comparison of the performance of single and two compartment septic tanks
Non-recirculating
septic tank effluent
(19 sites)
Overall
Mean
Overall
Median
Mean of
independent
averages
Median of
independent
averages
Geom.
Mean
Oregon
Rule Res.
Waste
Strength
Percent of
samples
exceeding
the rule Count
BOD-5 (mg/l) 264 250 246 242 236 300 32% 379
Total Suspended
Solids 85 61 79 66 64 150 10% 378
Total nitrogen 65 63 63 64 63 150 1% 377
Oil & grease 35 28 32 29 29 25 59% 367
Single
compartment
septic tank effluent
(9 sites)
Overall
Mean
Overall
Median
Mean of
independent
averages
Median of
independent
averages
Geom.
Mean
Oregon
Rule Res.
Waste
Strength
Percent of
samples
exceeding
the rule Count
BOD-5 (mg/l) 263 240 260 249 236 300 32% 218
Total Suspended
Solids 103 71 103 82 78 150 17% 218
Total nitrogen 63 62 63 62 61 150 0% 216
Oil & grease 32 27 33 29 29 25 56% 211
Two compartment
septic tank effluent
(10 sites)
Overall
Mean
Overall
Median
Mean of
independent
averages
Median of
independent
averages
Geom.
Mean
Oregon
Rule Res.
Waste
Strength
Percent of
samples
exceeding
the rule Count
BOD-5 (mg/l) 250 240 246 235 235 300 28% 148
Total Suspended
Solids 59 49 57 51 51 150 2% 147
Total nitrogen 66 64 66 65 65 150 1% 148
Oil & grease 30 28 32 28 29 25 61% 143
The table includes the mean and median values for the total population, the averages of the
individual means and medians for each tank and the geometric mean. The geometric mean may be a
useful tool in this circumstance as the data for the individual tanks is slightly skewed. The geometric
mean is also useful to reduce the effect of varying sample sizes. Converse (2004) applied this
method in order to better compare data from sample sets with significantly different sample sizes
(between 31 and 517). The sample size effect is lessened in this study because the counts are much
closer in value but the geometric mean also provides this benefit.
From this data, it appears that the two-compartment septic tanks perform better than single
compartment tanks on total suspended solids even when considering the median or geometric mean
values. The BOD-5 reduction is slightly better but the difference is not statistically significant. A
large number of septic tanks serve residences where at least one person in the household is taking
prescription medication for a long period of time (six two-compartment tanks and four one-
compartment tanks). When these tanks are eliminated from the analyses, the results are as shown in
Table 5. While these tanks perform better than the total population, the effluent still exceeds the
residential waste strength definition a statistically significant portion of the time.
Table 5. Comparison of the performance of single and two compartment septic tanks with known
prescription drug users removed.
Single compartment
septic tank effluent
(5 tanks)
Overall
Mean
Overall
Median
Mean of
Independent
Averages
Median of
Independent
Averages
Geometric
Mean
Percent of
samples
exceeding Count
BOD-5 (mg/l) 258 250 265 251 237 33% 122
Total Suspended
Solids 128 83 125 96 89 23% 122
Total nitrogen 64 63 64 63 62 0% 122
Oil & grease 33 29 35 33 32 61% 116
Two compartment
septic tank effluent
(4 tanks)
BOD-5 (mg/l) 234 250 235 228 226 23% 64
Total Suspended
Solids 61 51 61 53 56 2% 64
Total nitrogen 74 65 75 74 73 3% 64
Oil & grease 21 19 21 20 19 30% 63
The only parameter for which the residential waste strength definition appears to be valid is for total
nitrogen or total Kjeldahl nitrogen. This result was somewhat of a surprise because one hypothesis
was that the waste strength was greater because of water conserving fixtures. If that were the case,
however, the TN results would be much higher because of the lack of dilution. It appears that the
waste strength may be more strongly influenced by the chemically or biologically active inputs to
the systems than water conserving fixtures. Also, given that ten of the nineteen non-recirculating
septic tanks included in the project include residents taking some kind of long-term prescription drug
(four are families with children or young couples and six are retired couples), chemically or
biologically active inputs may be present in a significant proportion of the population.
The project team also considered how the results might differ between sample locations, in particular
what differences might exist between free-flowing grab samples taken from the discharge pipe of the
treatment unit and a grab sample taken from the pump chamber following the treatment unit. To test
the hypothesis, we took a series of side-by-side samples from the effluent pipe of three different
types of treatment units and the pump chamber following the treatment unit for a total of six systems
over a six to ten month period. The sampling team collected the effluent pipe sample by waiting for
the normal discharge cycle to occur. The chamber sample was collected by dipping the sample from
the pump chamber to which the effluent pipe discharged. The data, including correlation
coefficients, are presented in Table 6 and data plots are provided in Figures 2 through 19.
Table 6. Correlation coefficients for comparative sampling from the effluent pipe vs. the collection
chamber.
System type System name BOD-5 TSS TN Fecal E. coli
Retention time in
collection chamber
RX-30 System-R 0.99 0.64 0.99 0.91 0.94 0.30 day
RX-30 System-H2 0.22 0.72 0.99 0.96 0.09 0.24 day
RX-30 System-M 0.64 -0.20 0.79 -0.21 -0.23 0.67 day
NITREX System-F 0.99 0.92 0.99 0.06 0.01 N/A
NITREX System-S 0.98 0.61 0.99 0.48 0.28 N/A
FAST System-P 0.89 0.98 0.97 0.98 0.98 N/A
During this process the question also arose of the possible compositing effects of the septic tank,
treatment unit and other components of onsite wastewater treatment systems. In most of the single-
pass septic tanks, the household daily flow ranged between 4% and 20% of the tank volume with the
average flow being 10% of the tank volume. Potentially, this is a composite, or at least a blending,
of the number of days that is equivalent to the residence time in the tank or chamber. Unfortunately,
the project team could not warrant the extra expense and effort required to use composite samplers to
test the hypothesis that the septic tank and treatment unit (if present) are compositing devices. This
could be a valuable avenue of investigation in the future because if it proves to be true it would
maximize the limited resources of homeowners, service providers, or local jurisdictions to perform
sampling on onsite systems.
The correlation coefficients shown in Table 6 provide an overview of how representative effluent is
when sampled from the collection chamber following the treatment versus a grab sample taken from
the treatment unit effluent pipe. In most cases, the collection chamber sample quality correlates
highly with the effluent pipe sample for BOD-5. Two exceptions, the System-M and System-H2
RX-30 units show discrepancies in the results. The field observation records show significant upsets
in the operation of System-M during this sample period. The records for System-H2 show some
non-specific issues (solids sloughing from textile filter) that suggest operational upsets with this
system as well. Apart from these issues, it appears that the most consistent results between the two
sample locations are the BOD-5 and TN. The total nitrogen pipe vs. chamber sampling results are
very highly correlated with the one poor correlation produced by the system with numerous
operational upsets.
Total suspended solids pipe and chamber results are not as well correlated with each other. This
may be due to difficulties in sampling the chamber without disturbing the attached growth on the
chamber walls and pump fittings in those cases where the TSS is higher in the chamber than from
the pipe. The instances where the results from the pipe are higher may indicate settling in the pump
chamber.
The bacteria pipe and chamber results do not correlate well at all and are highly variable across the
six systems. The project team expected that the bacteria results would not be representative when
sampling from the chamber because of either die off or growth, depending on the conditions in the
chamber. Additionally, standard operating procedures for wastewater sampling require that bacteria
samples be taken as grab samples. (Sams [Bush], 2003)
Figures 2 through 7 show the BOD-5 data for all six systems. A visual inspection of the data
illustrates how well the parameter correlates over time. Again, the data for Systems H2 and M
indicate the difficulties these systems experienced during the sampling period. Generally, however,
the data from the chamber follows the trend of the pipe discharge in both cases, which implies that
the chamber data is a good relative indicator of performance. Systems-R, -F, and –S show such high
correlations and strong relationship over time that there is no practical difference between sampling
the effluent pipe or pump chamber. The project team believes that this is due to what is a blend of
several days household effluent leaving the septic tank. This effluent is further blended by treatment
in the unit following the septic tank. As stated previously, the extent to which this blending can be
relied upon to produce a representative sample is a potentially valuable avenue of investigation.
Overall the systems, it is difficult to determine if the strength of the correlation is related to whether
or not the advanced treatment system recirculates. The RX-30 systems in this study recirculate and
the NITREX and FAST systems do not.
The TSS data is quite variable over the six systems. In general, the trend of the chamber data
follows that of the pipe data as shown in Figures 8 through 13. However, in several instances, there
are considerably higher levels of TSS in the chamber vs. the pipe. This may be due to sampling
error if the sampler dislodged attached growth from the walls of the chamber or the pump fittings or
suspended growth if the effluent in the chamber is of such a quality as to support secondary growth.
There are also several instances when there is greater TSS in the pipe sample than the chamber. This
could be due to settling of solids in the pump chamber. An example of this is shown in Figure 10
where the pipe effluent is higher than the pump chamber effluent by approximately 10 mg/L. From
the chart it appears that this is a significant difference from the quality of the effluent taken from the
pump chamber. The negative correlation of the data points to a poor relationship between the date
from the two locations. However, when the first three months data is removed from the data set, the
correlation is good for the remainder of the sample record (r = 0.77). Given this, the project team
felt that the TSS results from the chamber were not quite as representative of the quality of effluent
that the advanced treatment unit produced but still good.
System-R BOD5 Chamber vs. Pipe over time
0
2
4
6
8
10
12
14
16
18
20
11/1
3/20
00
12/1
3/20
00
1/13
/200
1
2/13
/200
1
3/13
/200
1
4/13
/200
1
5/13
/200
1
mg
/L
Chamber BOD5
Pipe BOD5System-R BOD5 scatter plot
0
2
4
6
8
10
12
14
16
18
0 2 4 6 8 10 12 14 16 18 20
Chamber (mg/L)P
ipe (
mg/L
) P BOD5
Linear (P BOD5)
r = 0.99
r2 = 0.98
Figure 2. RX-30 System-R Pipe vs. Chamber BOD-5 Data
System-H2 BOD5 Chamber vs. Pipe over time
0
5
10
15
20
25
11/1
5/20
00
12/1
5/20
00
1/15
/200
1
2/15
/200
1
3/15
/200
1
4/15
/200
1
5/15
/200
1
mg
/L
Chamber BOD5
Pipe BOD5
System-H2 BOD5 scatter plot
0
5
10
15
20
25
0 2 4 6 8 10 12 14 16 18
Chamber (mg/L)
Pip
e (
mg
/L)
r = 0.22
r2 = 0.05
Figure 3. RX-30 System-H2 Pipe vs. Chamber BOD-5 Data
r = 0.99 r2 = 0.98
System-M BOD5 Chamber vs Pipe over time
0
2
4
6
8
10
12
14
16
12/1
2/20
00
1/12
/200
1
2/12
/200
1
3/12
/200
1
4/12
/200
1
5/12
/200
1
6/12
/200
1
mg
/L
Chamber BOD5
Pipe BOD5
System-M BOD5 chamber vs. pipe
0
2
4
6
8
10
12
14
16
18
0 50 100 150 200 250
Chamber (mg/L)
Pip
e (
mg
/L)
BOD-5
Linear (BOD-5)
r = 0.64
r2 = 0.41
Figure 4. RX-30 System-M Pipe vs. Chamber BOD-5 Data
System-F BOD5 over time
0
50
100
150
200
250
300
12/26/2000 1/26/2001 2/26/2001 3/26/2001 4/26/2001 5/26/2001
mg
/L
Chamber BOD5
Pipe BOD5
System-F BOD5 scatter plot
0
50
100
150
200
250
300
0 50 100 150 200 250 300
Chamber (mg/L)
Pip
e (
mg
/L)
P BOD5
Linear (P BOD5)
r = 0.99
r2 = 0.98
Figure 5. NITREX System-F Pipe vs. Chamber BOD-5 Data
System-S BOD-5 over time
0
20
40
60
80
100
120
140
160
180
12/26/2000 1/26/2001 2/26/2001 3/26/2001 4/26/2001 5/26/2001
mg
/L
Chamber BOD5
Pipe BOD5
System-S BOD5 scatter plot
0
20
40
60
80
100
120
140
160
180
0 20 40 60 80 100 120 140 160 180
Chamber (mg/L)
Pip
e (
mg
/L)
P BOD5
Linear (P BOD5)
r = 0.98
r2 = 0.96
Figure 6. NITREX System-S Pipe vs. Chamber BOD-5 Data
System-P BOD5 chamber vs. pipe over time
0
5
10
15
20
25
30
35
40
45
2/7/
2001
3/7/
2001
4/7/
2001
5/7/
2001
mg
/L
Chamber BOD5
Pipe BOD5
System-P BOD Scatter plot
0
5
10
15
20
25
30
35
40
45
0 5 10 15 20 25 30 35
Chamber (mg/L)
Pip
e (
mg
/L)
P BOD5
Linear (P BOD5 )
r = 0.89
r2 = 0.79
Figure 7. FAST System-P Pipe vs. Chamber BOD-5 Data
System-R TSS chamber vs. pipe over time
0
1
2
3
4
5
6
7
8
9
10
11/1
3/20
00
12/1
3/20
00
1/13
/200
1
2/13
/200
1
3/13
/200
1
4/13
/200
1
5/13
/200
1
mg
/L
Chamber TSS
Pipe TSS
System-R TSS scatter plot
0
1
2
3
4
5
6
0 1 2 3 4 5 6 7Chamber (mg/L)
Pip
e (
mg
/L)
P TSS
Linear (P TSS)
r = 0.64
r2 = 0.41
Figure 8. RX-30 System-R Pipe vs. Chamber TSS Data
System-H2 TSS chamber vs. pipe over time
0
5
10
15
20
25
30
11/1
5/20
00
12/1
5/20
00
1/15
/200
1
2/15
/200
1
3/15
/200
1
4/15
/200
1
5/15
/200
1
mg
/L
Chamber TSS
Pipe TSS
System-H2 TSS scatter plot
0
2
4
6
8
10
12
14
16
18
20
0 5 10 15 20 25 30
Chamber (mg/L)
Pip
e (
mg
/L) P TSS
Linear (P TSS)
r = 0.72
r2 = 0.52
Figure 9. RX-30 System-H2 Pipe vs. Chamber TSS Data
System-M TSS Chamber vs. Pipe over time
0
10
20
30
40
50
60
70
11/1
3/00
12/1
3/00
1/13
/01
2/13
/01
3/13
/01
4/13
/01
5/13
/01
6/13
/01
mg
/L
Chamber TSS
Pipe TSS
System-M TSS Scatter Plot
0
2
4
6
8
10
12
14
16
18
20
0 10 20 30 40 50 60 70
Chamber (mg/L)
Pip
e (
mg
/L)
P TSS
Linear (P TSS)
r = -0.20
r2 = 0.04
Figure 10. RX-30 System-M Pipe vs. Chamber TSS Data
System-F TSS over time
0
5
10
15
20
25
30
35
40
45
50
12/26/2000 1/26/2001 2/26/2001 3/26/2001 4/26/2001 5/26/2001
mg
/L
Chamber TSS
Pipe TSS
System-F TSS scatter plot
0
2
4
6
8
10
12
14
0 5 10 15 20 25 30 35 40 45 50
Chamber (mg/L)
Pip
e (
mg
/L)
P TSS
Linear (P TSS)
r = 0.92
r2 = 0.85
Figure 11. NITREX System-F Pipe vs. Chamber TSS Data
System-S TSS over time
0
2
4
6
8
10
12
14
16
18
12/26/2000 1/26/2001 2/26/2001 3/26/2001 4/26/2001 5/26/2001
mg
/L
Chamber TSS
Pipe TSS
System-S TSS scatter plot
0
2
4
6
8
10
12
14
0 2 4 6 8 10 12 14 16 18
Chamber (mg/L)
Pip
e (
mg
/L)
P TSS
Linear (P TSS)
r = 0.61
r2 = 0.37
Figure 12. NITREX System-S Pipe vs. Chamber TSS Data
System-P TSS chamber vs. pipe over time
0
5
10
15
20
25
30
2/7/2001 3/7/2001 4/7/2001 5/7/2001
mg
/L
Chamber TSS
Pipe TSS
System-P TSS scatter plot
0
5
10
15
20
25
30
0 5 10 15 20 25Chamber (mg/L)
Pip
e (
mg
/L)
P TSS
Linear (P TSS)
r = 0.98
r2 = 0.96
Figure 13. FAST System-P Pipe vs. Chamber TSS Data
The total nitrogen data provided in Figures 14 through 19 show the highest correlations of the study.
In each case, the nitrogen data from the pump chamber correlated very strongly with the effluent
pipe data regardless of the systems’ performance or operational history. The lowest correlation of
the group, System-M at r = 0.79, is lower only because the first set of sampling results of the period
are quite different. While this r-value is quite a bit lower than the rest of the systems studied, it is
still considered a high correlation according to conventional statistical methods (Osborn, 2003).
These methods typically consider r>0.70 to be highly correlated. The correlation coefficient of the
remainder of the sampling period equals 0.96, in par with the other systems.
Given these results, the project team believes that total nitrogen data is representative of the systems’
performance and of the quality of effluent discharged to the dispersal field regardless of whether the
sample is taken from the effluent pipe or the pump chamber following the treatment unit. Originally,
the project team wondered if additional nitrogen reduction would occur in the pump chamber
because of the attached and suspended growth typically observed in pump chambers. However,
given the lack of difference between the results, this does not appear to be happening. Also, the
compositing nature of the upstream components may help minimize the potential differences
between the two sampling locations.
In general, the project team decided to take grab samples from the effluent pipe leaving the treatment
unit to represent the performance of that unit whenever possible and at least for the bacteria samples
if the total sample suite could not be taken from free falling effluent.
System-R TN Chamber vs. Pipe over time
0
5
10
15
20
25
30
35
40
11/1
3/20
00
12/1
3/20
00
1/13/
2001
2/13/
2001
3/13/
2001
4/13/
2001
5/13/
2001
mg
/L
Chamber TN
Pipe TN
System-R TN scatter plot
0
5
10
15
20
25
30
35
40
0 5 10 15 20 25 30 35 40Chamber (mg/L)
Pip
e (
mg
/L) P TN
Linear (P TN)
r = 0.99
r2 = 0.98
Figure 14. RX-30 System-R Pipe vs. Chamber Total Nitrogen Data
System-H2 TN chamber vs. pipe over time
0
5
10
15
20
25
30
11/1
5/20
00
12/1
5/20
00
1/15
/200
1
2/15
/200
1
3/15
/200
1
4/15
/200
1
5/15
/200
1
mg
/L
Chamber TN
Pipe TN
System-H2 TN scatter plot
0
5
10
15
20
25
30
5 7 9 11 13 15 17 19 21 23 25
Chamber (mg/L)
Pip
e (
mg
/L)
Total Nitrogen
Linear (Total Nitrogen)
r = 0.99
r2 = 0.98
Figure 15. RX-30 System-H2 Pipe vs. Chamber Total Nitrogen Data
System-M TN chamber vs. pipe
0
10
20
30
40
50
60
70
80
11/1
3/20
00
12/1
3/20
00
1/13
/200
1
2/13
/200
1
3/13
/200
1
4/13
/200
1
5/13
/200
1
6/13
/200
1
mg
/L
Chamber TN
Pipe TN
System -M TN Scatter Plot
0
10
20
30
40
50
60
20 30 40 50 60 70 80
Chamber (mg/L)
Pip
e (
mg
/L)
Total Nitrogen
Linear (Total Nitrogen)
r = -0.79
r2 = 0.62
Figure 16. RX-30 System-M Pipe vs. Chamber Total Nitrogen Data
System-F TN over time
0
5
10
15
20
25
30
35
40
45
12/26/2000 1/26/2001 2/26/2001 3/26/2001 4/26/2001 5/26/2001
mg
/L
Chamber TN
Pipe TN
System-F TN scatter plot
0
5
10
15
20
25
30
35
40
0 5 10 15 20 25 30 35 40 45Chamber (mg/L)
Pip
e (
mg
/L)
P TN
Linear (P TN)
r = 0.99
r2 = 0.98
Figure 17. NITREX System-F Pipe vs. Chamber Total Nitrogen Data
System-S TN over time
0
5
10
15
20
25
12/26/2000 1/26/2001 2/26/2001 3/26/2001 4/26/2001 5/26/2001
mg
/L
Chamber TN
Pipe TN
System-S TN Scatter plot
0
2
4
6
8
10
12
14
16
18
20
0 2 4 6 8 10 12 14 16 18 20
Chamber (mg/L)
Pip
e (
mg
/L)
P TN
Linear (P TN)
r = 0.99
r2 = 0.98
Figure 18. NITREX System-S Pipe vs. Chamber Total Nitrogen Data
System-P TN chamber vs pipe over time
0
10
20
30
40
50
60
70
2/7/2001 3/7/2001 4/7/2001 5/7/2001
mg
/L
Chamber TN
Pipe TN
System-P TN scatter plot
30
35
40
45
50
55
60
30 35 40 45 50 55 60 65
Chamber (mg/L)
Pip
e (
mg
/L)
P TN
Linear (P TN)
r = 0.97
r2 = 0.94
Figure 19. FAST System-P Pipe vs. Chamber Total Nitrogen Data
Conclusion
Sampling onsite wastewater treatment systems has its roots in the standard operating procedures for
sampling larger centralized systems but with some differences. Further examination is necessary to
determine the standard operating procedures for onsite systems that will support a long term
sustainable monitoring program for the local authority and private constituents. The suite of sample
parameters taken from a system will largely be driven by site-specific or regional concerns. The La
Pine Project, located in a nitrogen sensitive area, concentrated on BOD-5, TSS and the nitrogen
species as the primary constituents for evaluating systems performance. A secondary suite of
parameters plays a supporting role, including bacteria, total alkalinity, chloride, and the field
parameters of DO, pH, conductivity and temperature.
The type of sampling equipment can be an important factor in the ease and safety of sampling a
wastewater system. Consistency of technique and personnel reflects on the consistency and
reliability of the dataset. Consistency of the dataset also relies upon using consistent and
representative sampling stations. A grab sample from the pump chamber following the treatment
unit correlates strongly to the end of the pipe discharge for BOD-5 and total nitrogen and relatively
well for TSS. Therefore, if no other option is available, sampling from the pump chamber can be
substituted for the effluent pipe sample if necessary. Bacteria results from the pump chamber are
suspect and standard procedure should be to sample from the effluent pipe if at all possible.
The hypothesis that septic tanks, treatment units and pump chambers have a compositing effect on
the effluent needs to be tested with side by side comparative sampling. The potential financial
benefit to the homeowner, local permitting agency and maintenance service provider of an
alternative to composite sampling would be significant.
References
Burks, B.D. and M.M. Minnis. 1994. Onsite Wastewater Treatment Systems. Hogwarth House,
Limited, Madison, WI.
Converse, J.C. 2004. Effluent Quality From ATUs and Packed Bed Filters Receiving Domestic
Wastewater Under Field Conditions. In Proceedings of the 10th
National Symposium of Individual
and Small Community Sewage Systems. American Society of Agricultural Engineers, St. Joseph,
MI.
Crites, R. and G. Tchobanoglous, 1998. Small and Decentralized Wastewater Management Systems,
McGraw Hill, Boston, MA.
Oregon DEQ (Department of Environmental Quality), 2000. Oregon Administrative Rule, Chapter
340, Division 71, Onsite Sewage Disposal.
Osborne, J.W. 2003. Effect Sizes and the Disattenuation of Correlation and Regression Coefficients:
Lessons from Educational Psychology. In Practical Assessment, Research & Evaluation.
Downloaded from PAREonline.net, 9/1/2004.
Sam, G.L. (J. Bush). 2003. Effluent Sampling for Residential Treatment Systems. In Proceedings of
the 2003 Conference of the National Onsite Wastewater Recycling Association. NOWRA,
Edgewater, MD.