ARCTIC EVALUATION
OFA
SMALL PHYSICAL-CHEMICAL SEWAGE TREATMENT PLANT
U. S. ENVIRONMENTAL PROTECTION AGENCY
ARCTIC ENVIRONMENTAL RESEARCH LABORATORY
COLLEGE, ALASKA 99701
-------
-------
' A/,
ARCTIC EVALUATION
OF A
SMALL PHYSICAL-CHEMICAL SEWAGE TREATMENT PLANT
H. J. Coutts
Environmental Protection Agency
Arctic Environmental Research Laboratory
College, Alaska 99701
Working Paper No. 16
October 1972
HEADQUARTERS LIBRARY
ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
-------
A Working Paper presents results of investigations
which are to some extent limited or incomplete.
Therefore, conclusions or recommendations—expressed
or implied—are tentative. Mention of commercial
products and/or trade names does not constitute
indorsement.
-------
TABLE OF CONTENTS
INTRODUCTION
OPERATING EXPERIENCES
PROCESS PERFORMANCE
OPERATING COSTS
CONCLUSIONS
APPENDIX
PAGE
1
2
10
14
16
17
-------
INTRODUCTION
In Arctic Alaska there are many transient industrial camps of 10-60
men each. Oil well drilling camps on the Arctic North Slope are usually
set up at one site for no more than one to two months. These camps contain
about 30 men each. The extended aeration package plant process is the
most common form of secondary waste treatment used. The major disad-
vantage of such units is that they require skilled operators and from
two to ten weeks to build up an operating biomass before efficiency can
be achieved.
The more concerned camp managers in the Arctic recognize the problems
with the extended aeration units and have been seeking other means of
providing secondary or higher level sewage treatment for their temporary
camps. Physical-chemical treatment is one method being considered to
guarantee secondary and in some cases achieve tertiary treatment. The
physical-chemical process involves chemical clarification followed by
carbon adsorption.
One such plant, a prototype installation, was operated consecutively
at two different sites near Prudhoe Bay in April and in May of 1972.
This paper will discuss the operation and performance of the unit.
Initially, this 7,000 gpd physical-chemical (P-C) unit was set up in late
March at the Nabors Drilling Co. #1-3 Drill Site east of the Deadhorse
airport and operated for approximately nne month. Three sets of samples
from April 12 to 17 were obtained from this site. The unit was then set
up at the Arco #1 Drill Site (approximately one mile from the Arco
Prudhoe Bay base camp) and operated during the month of May. Seven sets
of samples were obtained from that site.
1
-------
OPERATING EXPERIENCES
The block flow diagram of the unit is shown in Figure 1 (Met-Pro
Drawing Number 0003902): raw sewage is collected and pumped from a sump
below the housing trailer floor into an approximately 3500 gallon aerated
equalization tank. From the equalization tank the sewage and coagulant
are separately pumped into a flash mixer. From the flash mixer flow is
by gravity into the flocculation (center) section of an upflow clarifier.
Effluent (supernatant) from the clarifier then flows along with a small
airstream (for fluidization and to prevent anaerobiosis) up through a
carbon adsorption column (upflow) and into an 80 gallon surge tank. Liquid
from the surge tank is pumped through a pressure sand filter. Effluent
from tms filter is chlorinated by use of hypochlorite tablets and inter-
mittently siphoned onto the tundra.
The chlorinator consisted of a small dissolution box into which
four 3-inch slotted pipes {hypochlorite tablet containers) are standing.
Effluent flowing around and through the slots leaches hypochlorite from
the tablets. The baffled chlorine chamber has a theoretical design
detention of 1/2 hour.
Sludge from the bottom of the clarifier is pumped onto a moving paper
filter. Filtrate is pumped to the flash mixer. Sludge solids collected
on the filter paper are incinerated with the paper.
When activated by the high liquid level switch in the feed equalizer
tank the unit operates continuously at about 5 gpm design capacity until
shut off by a low level switch. This P-C process operated only at 100
percent of design capacity; the equalization tank absorbed sewage flow
fluctuations.
2
-------
r
solids to
incinerator
t RAW WASTE PUMP
Z FLASH MIX TAN*
3 COAGULANT FEEDER
4 DISINFECTANT FEEDER
5 FLOCCULATOR CLARIFIES
6 SLUDGE PUMP
7 DISPOSABLE MEDIA FILTER (OPT)
s ADSORBED
9 ADSORBER AERATOR
10 SURGE TANK
11 FILTER/BACKWASH PUMP
12 PRESSURE FILTER
Ol
^
o
C.J
6
ol
Figure 1
Process Flow Isometric
MET-PRO WATER TREATMENT CORP.
LANSDALE. PA. 19446
mLE SERIES 14000
I PC WASTEWATER TREATMENT
-------
Some process design parameters were:
Flash mix tank: ^3 gallon tank with a 1750 rpm propeller mixer.
Clarifier: Circular upflow with conical bottom and center-
mounted flocculator.
Overflow rate: 0.45 gpm/sq ft.
Sludge drawoff: Continuous at M/2 gpm.
Carbon Column: 24 inch diameter upflow - air fluidized.
Charge: 300 Ib hard (coal base) granular activated
carbon.
Contact time: 20± minutes.
Sand filter: Pressure downflow, single media.
Flow rate: 6 gpm/sq ft.
Paper filter: Adjusted to renew paper at ^6 inch water head.
The P-C unit was installed (by Steel Fabricators, Anchorage, Alaska)
along with the chlorine contact tank, the feed equilizer tank, sewage
sump and incinerator in a skid-mounted trailer, approximately 10x40 feet,
so that the whole waste handling system could be loaded into a C-130
aircraft (Here) and flown to the North Slope. Placement of the unit in
the trailer is as shown in Figure 2 and Figure 3 (Modified Met-Pro Drawing
Number 0003815). The disadvantage to such a compact enclosure was that
there was too little access room for easy operation and maintenance.
The only access to the unit with more than one foot clearance from any
wall was the side with the control panel. Clearance on that side was
limited to less than two feet.
As noted on Figure 3 access to the surface of the clarifier was
severely limited by the domed trailer roof; so limited, in fact, that
there was not enough room for the mixer to be installed on the flash-
mix tank.
The following problems were observed with the units during the
inspection trips. This listing is not meant to be a criticism but should
4
-------
o
X
5 o
o <—
—I X
x 2
LU O.
ft £
« CO
1_
= z
M i—«
s-
-------
DOMED ROOF ONLY
FIGURE 3
P-C UNIT EQUIPT. LAYOUT
-------
be used as a guide to avoid the problems in future installations involving
package P-C sewage treatment systems.
April 1972
Operational Problems
Excess coagulant (ferric sulfate)
in effluent.
High pressure drop through the
carbon adsorption column.
Air-fluidizing line clogged
with carbon particles.
Filter paper supply exhausted,
Contaminated air in trailer;
chlorine, smoke, and aerosols.
Excess noise in housing
trailer.
Sludge level sensor activated
by surface scum rather than
sludge blanket.
Probable Cause—-Suggested Cure
Operator not familiar with jar test.
Instruction manual should direct the
operator to observe the clarifier over-
flow for coagulant color and/or clarity.
Initial carbon charge too soft.
manufacturer specified carbon.
Use
Carbon particles too fine? Maintain
integrity of diffuser screen over air
outlet.
Lack of adequate supplies. Innovative
operator had to use rolls of paper towels,
Solid tablet dissolution system effer-
vescing chlorine gas. Use fewer tablets
and cover contact chamber.
Incinerator smoke coming through firewall
leak. Seal firewall.
Aerosols from aerated equalization tank.
Reduce aeration rate and/or cover tank.
Install inlet filter silencer on air
compressors.
Operator did not realize existence of
dimmer switch which would reduce
sensitivity. This was not discussed
in the instruction manual.
Operational Problems
Copper and plastic process line
breakage.
May 1972
Probabl e Cause—Suggested Cure
Moving units from Nabors to Arco pad.
Provide 100 percent spare for vulnerable
lines or fittings or change piping
materials to eliminate vulnerability.
-------
Operational Problems
Sludge filtrate return pump
burned out.
Relay switches in control
panel tripping (overheating).
Sewage solids clogging 3/4
inch (suction) feed line.
Low voltage turbidity sensing
light had been removed.
Chemical feeder pump, head
needed rebuilding (diaphragm).
Excessive foam over-flowing
equalizer tank.
Loss of media in the sand
filter.
Probable Cause--Suggested Cure
Unknown.
Electrical. Unknown. Feed pump line
clogging?
Use comminuter or easily cleaned suction
cage, or larger size pipe.
Short in light transformer? Transformer
replaced.
Grit in ferric sulfate. Premix ferric
sulfate and decant into chemical feeder
supply tank.
Too much aeration. Cycle air compressor
or reduce aeration rate. Install foam
control spray header.
Lost when thawing and removing ice from
unit. Drain all lines and tanks during
shutdown.
It should be noted that almost any new piece of process equipment
usually has a considerable number of operational problems during its
start-up and initial operational phases. Untrained and inexperienced
operators usually compound the problem. These problems were listed only
as a guide for considerations of future installations of similar units.
It should be remembered that this P-C unit was hastily installed in the
trailer. A less rushed job of preparation may have eliminated many of
the problems which were encountered. Many of the problems were minor
(none were unsurmountable) and the unit was operable at design capacity
after recharging with the proper activated carbon before operation at
the Arco site (May). The manufacturer-specified activated carbon (hard
coal base) was not available at Anchorage when the unit was to be shipped
to the site so the only available activated carbon (lignite base)
8
-------
substituted. Attrition of this charcoal during handling and operation
decreased its mesh size to the extent that it severely reduced capactiy.
Operators of P-C units in the Arctic seem to prefer ferric sulfate
as coagulant. The advantages of ferric salts as compared to lime and
alum are that: it operates over a wide concentration range, it dissolves
easier, effluent neutralization is usually not required, and a rust
clouded effluent may indicate its use in excess.
One disadvantage of P-C units is that they contain much intricate
machinery and it takes the operator a while before he can become familiar
with all the equipment; but he does not have to understand a biological
process as a good operator of an extended aeration unit must.
-------
PROCESS PERFORMANCE
At the Nabor's camp (April 1972) the unit treated an estimated 1,000
to 2,000 gpd of bathroom and personal laundry wastewater from about 33
men. Linens and towels were sent out for laundry. Kitchen wastes were
separately sewered and discharged into a lagoon.
After recharging with the proper size carbon (Calgon Filtrasorb 300}
the unit was set up at the Arco site on approximately May 2. All domestic
wastewater including kitchen wastes from the 33-man camp was treated in
the unit.
At both sites the feed (raw sewage) samples were collected as a small
slip stream from the sewage pump that discharges into the feed equalization
tank. A pipe Tee was attached to the end of this pump's 2-inch discharge
hose, and from that Tee a 3/8 inch sample hose was run into a 15 gallon
collection barrel. For a 24-hour composite the barrel was sampled and
emptied daily. When the 3/8 inch line plugged, grab samples were taken
from the equalization tank. The clarifier overflow samples were dipped
(all were grab samples) from the surface adjacent to the overflow weir.
Effluent samples were collected (in a 15 gallon barrel) as a slip stream
from the sand filter; but before chlorination. Both 15 gallon collection
barrels were set on the cold trailer floor in lieu of refrigeration.
An hour meter was attached to the positive displacement feed pump
and measurements taken from May 16 to May 23. The unit ran at 81 percent
of design capacity during one eleven-hour period (0800 to 1900) on May 17.
The average flow rate was 4,100 gpd; 58 percent of design capacity. The
per capita daily consumption averaged 120 gallons.
-------
Apparently in thawing the unit (before moving to Arco pad) the
operators inadvertently removed most of the media from the sand filter.
The effect of insufficient effluent filtration shows in the data for
May (Appendix), The most important parameters are summarized in Table 1
TABLE 1
Data Summary
Physical-Chemical Unit on the North Slope, 1972
FEED; Waste Water Source
Rate
COD mg/1
Total Nitrogen mg/1
Total Phosphorus mg/1
(Number of samples)
Clarifier Overflow:
COD mg/1
Suspended Solids mg/1
(Number of samples)
Effluent:
COD mg/1
Suspended Solids mg/1
Total Nitrogen mg/1
Total Phosphorus mg/1
(Number of samples)
April Average
Nabors Site
Bathroom and
Personal Laundry
Est. 1000-2000 gpd
1257
127
9.3
(3)
178
40
(3)
61
26
127
0.7
(3)
May Average
Arco Site
All Domestic
Camp Sewage
4100 gpd
846
48
15
(7)
439
119
(4)
312
155
32
3.4
(7)
The April data for the chemical precipitation-clarification step had
a COD removal ranging from 79 to 89 percent. The carbon adsorption-sand
filtration steps had a removal range from 61 to 71 percent. Average over-
all COD removal was above 95 percent. This is much higher than one could
expect with conventional extended aeration practice where the upper limit
is about 80 percent. The only BOD data point for April (Appendix) shows
a removal approaching 99 percent.
11
-------
The lack of effluent filtration shows in the data for May, but
the chemical precipitation-clarification step also had a very poor COD
removal, averaging about 48 percent. The inclusion of kitchen wastewaters,
raising grease and detergent levels, may account for a minor part of the
reduction in coagulation-clarification performance. The sludge level
sensor light was out of service (see operational problems) during the
May sample collection period. Without the sensor light there is no
control of clarifier effluent turbidity /hich is suspected as being the
major cause of poor clarifier and carbon column performance. The carbon
adsorption-sand filtration steps had a low average COD removal of 29
percent. It should be recognized that computations based on clarifier
surface grab samples are not as reliable as those based on composite
samples. The average overall COD removal for the May samples is about
63 percent. Average BOD (Appendix) removal is about 71 percent which
is about the performance to be expected for an extended aeration process
under similar conditions.
For total solids the April data (Appendix) shows about 650 mg/1
removed. For May, only about 400 mg/1 was removed. Again for April the
effluent suspended solids averaged about 26 mg/1, whereas for May the
average was about 155 mg/1. Turbidity data (appendix) follow the same
trend.
This physical-chemical process had an inconsistent effect upon the
removal of total and ammoniacal nitrogens. Average removals varied from
0 to 33 percent. P-C processes usually remove only organic nitrogen.
The effluent total phosphorus concentration could be kept below 1 mg/1
independent of feed concentration when enough coagulant is used and
suspended solids removal is effective. May data shows an average removal
of 77 percent. ,
-------
There was excellent total coliform removal (Appendix) before chlorin-
ation. With all four tablet dispensers in the chlorinator the effluent
total chlorine (0-tolidine method) concentration exceeded 5 ppm after a
theoretical contact time of 1/2 hour. The pH ranged from 6.3 to 7.9 which
indicates that in this case effluent neutralization was not required.
TABLE 2
Filter Paper Performance
Sampl e
Date
4/12/72
5/17/72
5/19/72
5/23/72
% Water in
Captured
Sludge
^90
91
95
98
Dry Solids
Capture
gm/in^ paper
0.1*
0.36
0.04
0.02
Dry Solids
% Volatile
49
49
49
49
Fe2(S04)3
Feed
Cone, mg/1
%300
^380
^380
^380
The performance of the paper filter is shown in Table 2. Only the
clarifier sludge was filtered. The filter system was set to operate
(expose fresh filter paper) when the pressure head on the paper exceeded
about 6 inches of water. The mass of the solids captured on the paper
varied over one order of magnitude.
The instruction manual stated that a scale (coagulant pump stroke
setting) of approximately 25 percent represented a feed strength of
approximately 250 ppm if coagulant was mixed 16 oz per gal. of water.
At the Arco site ferric sulfate was mixed 12 oz per gal. of water and
the scale set at 50 percent. The bulk density of the ferric sulfate is
about 70 Ib/cubic ft; for the above calculations it was assumed to be the
same as water. The usage rate averaged 15 Ibs per day since there
appeared to be considerable grit left after decanting the concentrated
ferric sulfate into the chemical feeder tank.
13
-------
OPERATING COSTS
Based upon flow and supplies information obtained at the Arco site
(except for carbon usage) the following cost estimates {materials: FOB
Anchorage) are made:
1. Manpower. The unit required about 6 ± 3 hours per day to operate.
This would represent in the order of about 1,0 cent per gallon but this
cost should not be directly attributed to the treatment since the operators
were untrained and unfamiliar with some of the equipment. After the
equipment was completely shaken down, adjusted and properly maintained,
the manpower requirement should be less than 3 hours per day.
2. The ferric sulfate cost at the dosages used was about 0.06 cents
per gallon of wastewater.
3. The filter paper usage varied from about 50 to 100+ yards per day.
Using an average of about 80 yards per day the filter paper charge would
then be about 0.15 rents per gallon wastewater treated.
4. A 300 Ib charge of activated carbon was used. Assuming a capacity
of 1 Ib of COD per Ib of activated carbon and an average clarifier overflow
COD of 180 mg/1 (Nabors site) the charge for replacing the activated carbon
would then be 0.07 cents per gallon of wastewater treated. Under conditions
as they were at the Arco site the carbon charge should last about 1 to 2
months.
5. Since the hypodilorite tablets were at times used in excess,
consumption figures were not recorded. The chlorination charge should
be less than 0.02 cents per gallon and was therefore neglected.
14
-------
Excluding manpower or equipment amortization the total operating
expense has been in the order of about 0.3 cents per gallon. Substitution
of an economical solids concentrator along with other equipment improve-
ments and operator training could substantially reduce operating costs.
15
-------
CONCLUSIONS
The performance characterizing this plant is based on sampling
periods when the process was not operating as intended. Therefore the
data does not show the actual capability of the process.
The process performed well attaining an overall COD removal above
95 percent in April, in spite of many operational problems. In May the
unit attained a secondary treatment level.
Considerable effort, for isolated arctic installations, needs to be
devoted towards proper design, set up, hardware selection and operation
if P-C units are to consistently provide a tertiary treatment level.
16
-------
APPENDIX
17
-------
GLOSSARY
Unless otherwise noted all analyses were performed in accordance
with EPA Methods for Chemical Analysis of Mater and Wastes,1971, and/or
Standard Methods - Mater and Mastewater, 13th Edition, 1971.
BOD - Biochemical oxygen demand expressed as elemental oxygen.
COD - Chemical oxygen demand expressed as elemental oxygen.
FILT COD - COD of filtrate through Gelman #61633 glass fiber filter.
TOC - Total organic carbon expressed as elemental carbon.
TS - Total solids.
TVS - Total volatile solids.
SS - Suspended solids.
VSS - Volatile suspended solids.
T. coli/100 ml - coliform bacteria enumeration using membrane filter
technique - counts per 100 milliliters.
TKN-N - Total Kjeldahl nitrogen expressed as elemental nitrogen.
NH^-N - Ammonical nitrogen expressed as elemental nitrogen.
T-PO^P - Total phosphate expressed as elemental phosphorus.
Fe - Iron.
Grease - grease and oil liquid-liquid extraction with chloroform.
mg/1 - milligrams per liter.
gpd - gallons per day.
pH - indicates acid <7.0 or base >7.0; 7.0 is neutral.
COLOR - True, absorbance at 380 millimicron wavelength.
TURBIDITY - Jackson turbidity units, Hach #2100 Turbidimeter.
18
-------
RESULTS OF ANALYSIS OF SAMPLES FROM NABORS SITE
Date
Location
4/12/72 4/14/72
Feed Clarif. Eff. Feed Clarif. Eff.
Surface Surface
4/17/72
Feed Clarif. Eff.
Surface
24-hr Composite? Yes
No
Yes Yes
No
No
No
No
Yes
COD mg/1
Filt. COD mg/1
TOC mg/1
BOD mg/1
TS mg/1
TVS mg/1
SS mg/1
VSS mg/1
TKN mg/1 N
NHo mg/1 N
T-P04 mg/1 P
T. coli/100 ml
PH
Color
Turbidity
1050
-
375
-
2000
940
840
470
110
38
13
-
7.5
-
-
223
-
93
-
1300
270
49
27
-
-
-
-
6.8
-
52
65 1450 158
_
26
_
1200
220
8
5
113
92
0.8
_
7.5
108
13
62 1270
-
-
490
- 1700
900
830
560
144
99
5.6
- 4x1 O8
8.1
_
160
153
138
-
-
1300
310
31
13
-
-
-
-
-
_
-
55
52
-
6
1200
210
43
13
140
107
0.6
<18
7.9
_
43
19
-------
RESULTS OF ANALYSIS OF SAMPLES FROM ARCO SITE
Date
Location
24- hr Composite?
COD mg/1
Filt. COD mg/1
TOC mg/1
BOD mg/1
TS mg/1
TVS mg/1
SS mg/1
VSS mg/1
TKN mg/1 N
NH3 mg/1 N
T-P04 mg/1 P
T. coli/100 ml
PH
Color
Turbidity
Fe mg/1
Feed
No
838
479
-
180
1800
710
370
220
53
45
25
5x1 08
7.6
283
-
7
5/17/72
Clarif ,
Surface
No
399
-
-
-
1500
330
190
100
36
30
3_
9x1 07
-
252
125
25
Eff.
Yes
134
60
-
33
1600
210
150
67
28
31
0.7
<20
6.3
139
130
52
5/18/72
Feed Clarif.
Surface
Yes
925
470
-
505
1900
760
530
290
19
17
55
5x1 09
7.3
307
-
-
No
179
-
-
-
1500
230
46
26
21
20
0.1
-
-
66
51
17
Eff.
Yes
306
155
-
105
1400
330
270
130
39
32
2.5
<10
6.5
114
125
48
Feed
Yes
701
498
-
227
1600
590
110
77
63
54
25
1.5x10
7.0
138
79
3
5/19/72
Clarif. Eff.
Surface
No
554
416
-
-
1500
520
no
70
58
44
a 20
i8 -
-
236
94
11
Yes
408
318
-
138
1400
420
48
31
51
36
14
<20
7.0
171
70
9
Date
Location
24-hr Composite?
COD mg/1
Filt. COD mg/1
TOC mg/1
BOD mg/1
TS mg/1
TVS mg/1
SS mg/1
VSS mg/1
TKN mg/1 N
NH3 mg/1 N
T-P04 mg/1 P
T. coli/100 ml
PH
Color
Turbidity
Fe mg/1
5/20/72 5/21/72 5/22/72
Feed Eff. Feed Eff. Feed Eff.
No
700
463
172
-
-
-
-
-
60
41
25
-
7.7
-
-
-
No No
1 59 848
95
49 238
-
-
-
_
-
14
11
0.7 -
_
6.8
-
-
-
No No
459 925
115
136 326
_
-
-
-
-
29
9.1
0.2
-
8.0
-
-
54
No
465
116
118
-
-
-
-
-
24
8
2
-
7.0
-
100
207
5/23/72
Feed Clarif. Eff.
Surface
No
986
311
313
360
2200
790
780
420
61
.1 40
.3 7.6
3x1 09
7.0
282
140
48
No
622
-
-
-
1600
530
130
68
-
_
_
-
6.4
265
110
-
Yes
250
169
76
78
1500
260
150
76
41
22
0.4
<10
6.3
65
110
41
20
. S. GOVERNMENT PRINTING OFFICE.- 1974-798-9*2/6 REGION 10
-------
------- |