OPERATION, LIMITATIONS, AND RECOVERY OF SAND FILTERS
William L Stuth[1]
Of
the twenty-nine sites studied in the Part I phase of this study, twenty-one
sites were found to be operating without ponding or surfacing. In Table 1, column 2, you will see that the
average age of the twenty-one sites is 3.5 years, the oldest site being 6 years
old.
The
third column reflects the use of time controllers. Each site should have had time controllers installed. The number "1" indicates that a
timer, approved for use in an on-site system, was installed and set
properly. Out of twenty-one sites 10
sites had the correct timers and the correct settings. Four sites had no time controllers and were
working as a demand system. These sites
are marked with a "0". The
remaining 7 sites had time controllers that were not approved for on-site
system use and are inconsistent, one approved timer had been turned to manual
by the owner, and the remaining timers were approved but set to operate in a
demand mode. This is accomplished by
reversing the values. As an example, a
system that by code must cycle 4 times within 24 hours, if set 6 hours on instead
of 6 hours off, will work as a demand system controlled by the floats. These sites are marked with an "1
/".
The
fourth column shows whether or not the sand collected at each site met the
specifications for C-33 sand. Sites
where the sand did not meet the specifications are marked with an
"X". All eleven of the
twenty-one sites where the sand failed the C-33 test, had sand that failed in
the 100 and 200 sieves with a higher percentage passing through than was
allowed, or more retained in the pan under the 200 sieve. This does not indicate however, that at the
time of installation, the sand would not have met the specifications. In the mid 1990s the sieve testing method to
determine C-33 sand changed from a dry sieve to a wet sieve. This change increased the amount of fines
retained in the 200 sieve. This is the
reason many of the samples failed. The
installer provided the original paper work for two of the sites and the sieve
reports that were provided by the gravel companies show that the sand did meet
the standards in force at that time. To
attempt to investigate each site for original paperwork would have been
extremely time consuming and costly.
Time and resources did not allow further follow-up.
Column
5 shows the average BOD5 for the twenty-one sites was 141 mg/L, with
a range from 216 mg/L to 26 mg/L. The
low BOD5 was collected at a site with a small cabin on the water,
and appeared to have very little use.
Groundwater infiltration was also very possible.
Column
6 shows the BOD5 of the effluent leaving the sand filters
and the in-ground mound. The average
BOD5 was 5.5 mg/L, with a range from 21.2 mg/L to 1 mg/L.
The
gallons per day (GPD) the system was designed to treat are shown in column
7. The number of bedrooms in the homes
determined all GPD. The average is 454
GPD, with a range from 480 GPD to 360 GPD.
Note: all sand filter systems
have the same design-loading rate of 1.2 gallons per square foot per day.
Column
8 shows the true GPD each site was using.
The average is 189 GPD, with a range from 446 GPD to 59 GPD. These flows were determined from hour meters
installed at each site or from water meters when available.
Total
BOD5 loading to the filter is shown in column 9. The product of true flow and BOD5
determine the total BOD5 load.
The average total BOD5 load was .20 lb/day; with a range from .65 lb/day to .02 lb/day.
Column
10 is the total BOD5 loading per square foot per day. In the past this element has been ignored,
but may be very valuable information when evaluating any on-site system. To determine the true load BOD5 per square foot you must first determine the
true flow of the system and the BOD5 the sand or soil is
receiving. True BOD5 loading
per square foot is the product of the GPD x BOD5 x .00000834 and
dividing the answer into the square foot area of the sand filter or other disposal
system. For example: assume
(1)
the GPD = 200
(2)
the BOD5 = 140
(3)
soil system or filter size = 375 square feet.
Then:
200
GPD x 140 BOD5 x .00000834 = .23 PPD and,
23
(PPD) ¸ 375 (ft2)= .0006. PPD/ft2.
This
gives you the BOD5 load per
square foot, per day to the soil system or the sand filter.
Column
11 is the true hydraulic loading per square foot. In this study, the average is .48 gallons per square foot per day
and the range is from 1.11 gallon per square foot per day to .14 gallons per square
foot per day. All 21 sites in this
study were designed at a 1.2 gal per square foot per day.
The
eight sand filter systems that were found surfacing and/or ponding are shown in
Table II. It shows the same values as
Table I with one addition, Column 2.
This shows that all eight systems were either failing or near
failure. Surfacing indicating failure
and ponding indicating near failure.
The
average age of the eight systems is again 3.5 years. The average age of the four systems that were ponding and
surfacing is four years while the average age of the systems ponding only is
three years old.
Column
3 indicates the use of time controllers.
Systems with approved time controllers set correctly are shown as
"1". Systems that had either
unapproved timers or approved timers set incorrectly are shown as "1
/" and "0" indicates timers were not used and the system was
operating as a demand system. Of the
eight systems, only three were found to have approved timer controllers set
properly.
Column
4 shows that only four of the eight systems passed the current C-33 sieve
test. The four failed in the 100 or 200
sieve indicating there were more fines present than allowed.
Column
5 shows the average sand filter influent had a BOD5 of 252 mg/L with
a range from 416 mg/L to 150 mg/L.
Column
6 shows the septic tank effluent, or sand filter influent BOD5. Site #26 was an in-ground mound without any
way to collect a sample. At site # 28
the flooded bed was leaking into the pump reservoir. This had a major impact on the BOD5 values. If you include this value in the average of
the seven sites the average BOD5 was 49.9. If you eliminate site # 28 and use the remaining six sites, the
average BOD5 is 14.7.
Another challenge to this column may be site #s 22 & 24. Both sites have surfacing sewage and there
is no way of determining what percent of the flow was surfacing or what percent
of the flow was actually passing through the filter. If you remove these two sites from the study it only leaves four
sites, 23, 25, 27 & 29 where 100% of the flow is passing through the
filter. The average BOD5 for
these four sites is 17.57 mg/L.
Column
8 shows the average design flows at 472 GPD.
Column
9 shows the true GPD for all sites averaged 220 GPD, with a range from 546 to
114 GPD.
Column
10 shows the total BOD5 load
for each site. The average BOD5
is .51 lb/day with a range from 1.67 lb/day to .15 lb/day.
Column
11 shows the total BOD5 per square foot. The average is .00127, with a range from .00410 pounds per day
per square foot to .00038.
Column
12 shows the true hydraulic load per square foot. The average is .55 gallons per square foot per day with a range
from 1.36 gallons per square foot per day to .28. Again all of these systems were designed with a loading rate of
1.2 gallons per square foot per day.
Originally the goal for the Part II study
was to include five sites. This proved
unexpectedly difficult. Members of the
on-site industry were contacted to obtain referrals for sites they felt would
benefit from the ATU installations.
Five sites were located, however only two sites became available, either
because the homeowners did not want to participate in the program or it was
felt that a simple repair would alleviate their problem. A third site was found when one site
included in the Part I study qualified, and the homeowner was willing to
participate. A fourth site was located
in an adjacent county.
The
first ATU installed was at a site that had a mound system as the disposal
component of the on-site system. The
homeowner had purchased the property one-year prior to the study. Aqua Test,
Inc. had been employed to inspect the system three to six months after the new
owners moved into the home and found that the mound was surfacing. Ponding was also found when the inspection
routine was performed. The septic tank
had just been pumped out so lab tests were not drawn at the time of the first
visit. A sample was collected at a later
date showed the BOD5 was 174 mg/L.
The owner did not fill out the questionnaire but reported that there
were three people in the household, herself, her husband, a child, and she was
a stay at home mother. She stated that
she did use a lot of cleaning products and that they had never lived in a home
serviced by a septic system before. She
agreed to participate in the Part II study, but after construction started she
would not allow an access port installed over the inlet of the septic
tank. She thought it would ruin the
looks of her lawn. Because of this, an
influent baseline sample was not collected at the time the ATU was
installed. Had the owner stated her
objection before construction was started, the site would have been eliminated
from the study. The day the ATU was
installed, there was three inches of ponding in the mound. During the nine follow-up site visits, zero
ponding was reported.
The
second site selected for an ATU installation also had a stressed mound. The ponding level in the mound was measured
as 3" and effluent had surfaced on one occasion. The homeowner was very cooperative and filled out the
questionnaire. This revealed that there
were seven people living in the home and the average flow was 325 GPD. On the day the ATU was installed there was
over three inches of ponding. Eight
follow-up visits recorded zero ponding.
The
third site selected incorporated a sand filter with a thick organic biomat
which resulted in the effluent ponding on the surface daily. The homeowner reported they had purchased
the home a little over a year prior.
There were three people living in the home and the average flow was 157
GPD. The owner also reported the system
had a bad odor daily. The sand filter
was ponded and surfacing at the time of the ATU installation. The first follow-up visit, seven days after
the ATU was installed found the sand filter was still ponded but there was no
surfacing. This site was selected late
in the program so there were only four follow-up visits and at the time of the
last visit there was four inches of ponding but no surfacing. Note: the normal practice for this ATU
installation is to pump out the ponded liquid in a sand filter at the time of
installation, but for the purpose of this study, the sand filter was left
untouched.
The
fourth ATU was installed at a home served by a failing sand filter. The owner reported the system was designed
to handle 360 GPD and they were using 188 GPD.
However, sewage had surfaced eight times in the past twelve months and
it was currently surfacing out of the top of the sand filter. One of the occupants of the home had been on
antibiotics for a very long time. On the day the ATU was installed, the sand
filter was surfacing. This sand filter
was not pumped out at the time of installation. Six days after installation of the ATU the system was still
ponded but not surfacing. This was still the condition through the completion
of the project. As this site was also
selected late into the study, there were only four follow-up visits.
This
study demonstrates that reducing the BOD5 loading to the interface can enhance
the performance of sand filters. Sand
filters and ATUs offer an additional option to professionals in the on-site
industry. The use of these alternatives
helps to render on-site systems as a permanent solution to waste water
management needs. The study of these
systems indicates that all of the options evaluated require regular
inspections, routine maintenance, and process monitoring. The true value of this study is to confirm
these three important requirements.