Environmental Protection Technology Series
MINIMIZATION OF WATER USE IN LEAFY
VEGETABLE WASHERS
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-77-135
July 1977
MINIMIZATION OF WATER USE
IN LEAFY VEGETABLE WASHERS
by
Malcolm E. Wright
Agricultural Engineering Department
and
Robert C. Hoehn
Civil Engineering Department
Virginia Polytechnic Institute and State University
Blacksburg, Virginia 24061
Grant No. S-802958
Project Officer
Harold W. Thompson
Industrial Pollution Control Division
Industrial Environmental Research Laboratory
Corvallis, Oregon 97330
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Industrial Environmental Research
Laboratory, Cincinnati, U.S. Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents necessarily reflect
the views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
ii
-------
FOREWORD
When energy and material resources are extracted, processed, converted,
and used, the related pollutional impacts on our environment and even our
health often require that new and increasingly more efficient pollution control
methods be used. The Industrial Environmental Research Laboratory - Cincinnati
(lERL-Ci) assists in developing and demonstrating new and improved methodo-
logies that will meet these needs both efficiently and economically.
This report covers the construction and evaluation of an improved
leafy greens vegetable washing system. This system consisted of two series
drum immersion washers, each with associated settling tanks and moving belt
screens. Wash water was used in a counter-current flow regime. Results
obtained when comparing the prototype process to current commercial washing
systems were encouraging. Significant reductions in wash water requirements
and wastewater generation were reported; as was an increase in cleaning
efficiency.
It appears that this process modification will become a building block
in the development of economically achievable waste management systems for
the leafy greens processing industry. As a result this report should be of
interest to processors of leafy greens, designers of processing facilities,
equipment manufacturers and environmental regulatory agencies.
Further information on this project can be obtained by contacting the
Food and Wood Products Branch of lERL-Ci.
David G. Stephan
Director
Industrial Environmental Research Laboratory
Cincinnati
iii
-------
ABSTRACT
This project was undertaken to construct and test an improved leafy
greens washing system employing water recirculation, to characterize the
quality of the wash water and waste stream, and to make comparisons to con-
ventional washers. The prototype system produced a cleaner product while
reducing water requirements and consolidating waste loads.
The prototype system consisted of two drum immersion washers in series,
each with associated settling tanks, filters, and water recirculation systems.
Construction was similar to conventional washers but with modifications to
improve removal of floating trash and increase hydraulic agitation of
product. Fresh water input was limited to that required to replace water
carried off by the product plus a small, overflow, effluent stream from the
system.
The prototype was tested in a commercial processing plant during the
fall and spring harvesting seasons, 1975-76. Sixty-seven metric tons of
collards, spinach, and turnip greens were processed through the prototype
in 52 hours of actual operating time. Conventional washers were monitored
for 27 hours (38 tons) for comparison. Insect and bacteria counts, COD,
TSS, VSS, and several other water and product parameters were measured at
predetermined times and locations. Data were obtained to predict expected
waste loads from the products processed.
Economic considerations indicate that the annual fixed costs of owning
the prototype system would be approximately $600 per year more than the costs
of owning a conventional system, of comparable capacity. Operating costs,
however, were $100/day less for the prototype than for the conventional
system in an example problem using conditions similar to those at the test
site. These results would, of course, vary considerably depending on local
utility rates and other operating costs.
This report was submitted in fulfillment of Grant No. S802958 by the
Virginia Polytechnic Institute and State University under the partial
sponsorship of the Environmental Protection Agency. This report covers the
period from May 1, 1974 to January 31, 1977, and work was completed as of
January 31, 1977.
iv
-------
CONTENTS
Foreword
Abstract
Figures
Tables
Abbreviations
Acknowledgment
1. Introduction ....................... i
2. Conclusions ........................ 3
3. Recommendations ...................... 5
4. Prototype Washer System .................. 7
Washer design .................... 7
Water flow instrumentation ............. 15
Installation and modifications ........... 19
5. Procedures ........................ 22
Overview ...................... 22
Specific procedures ................. 27
6. Results and Discussion .................. 34
Operating parameters ................ 34
Product quality parameters ............. 48
Water quality parameters .............. 52
Summary of waste production from washers ...... 62
Economic comparisons ................ 66
References ............................. 70
Appendices ............................. 72
A. Operating parameters data ................. 72
B. Product quality data ................... 80
C. Water quality data .................... 84
-------
FIGURES
Number page
1 Diagram of washing system showing water and product flow
patterns, sampling sites, and water flow meter locations ... 8
2 Overhead view of washing system adjacent to Exmore plant.
Rotary sand tumbler and conveyor into plant are at right
and foreground 9
3 Diagram of prototype washer 10
4 Paddle wheel showing expanded metal covering and spoked end
construction. Elevated nozzle banks are shown in fore-
ground 12
5 Exit conveyor of washer number 2 13
6 View of washer side drains in operation 14
7 Moving belt screen in operation 16
8 Moving belt screen and trash collector. Compressed air hose
for removing trash from belt is shown in foreground 17
9 Prototype settling tank 18
10 HS flume meter number 4 with water level recorder 20
11 HS flume meter number 5 with water level recorder 21
12 Schematic showing water and product sampling sites fox
comparative study of new vs. conventional leafy greens
washing systems at Exmore Foods, Exmore, Va 24
13 Water flow rates vs. operating time, trial 1, Fall, 1975,
when processing collards with prototype system. Refer to
Figure 1 for meter locations 37
14 Water flow rates vs. operating time, trial 1, Spring,
1976, when processing spinach with prototype system.
Refer to Figure 1 for meter locations 37
vi
-------
Number Page
15 Water overflow rates from conventional washers vs.
operating time, trial 1, Spring, 1976, when processing
spinach on the east line 37
16 Product flow rate vs. operating time, trial 2, Fall, 1975,
when processing collards with the prototype system .... 41
17 Accumulated product input vs. operating time, trial 2,
Fall, 1975, when processing collards with the prototype
system 41
18 Summation percentages vs. particle size for grit
accumulated in the prototype system sub-unit 1; trial
5, Fall, 1975, when processing spinach 47
19 Summation percentages vs. particle size for grit
accumulated in the prototype system sub-unit 2; trial
5, Fall, 1975, when processing spinach 47
20 Summation percentages vs. particle size for grit
samples taken from conventional Washer 1, East Line,
trial 2, Spring, 1976 when processing spinach 47
21 Grit (inorganic solids) on spinach vs. accumulated product
at three sites in prototype system, trial 1, Spring;
unwashed product (Site 1), product exiting first washer
(Site 3), product exiting second washer (Site 4) 51
22 Grit (inorganic solids) on spinach vs. accumulated product
at three sites in conventional system, trial 1, Spring;
unwashed product (Site 7), product exiting first washer
(Site 8), product exiting second washer (Site 10) .... 51
23 Grit (inorganic solids) on turnip greens vs. accumulated
product at three sites in prototype system, trial 6,
Spring; unwashed product (Site 1), product exiting first
washer (Site 3), product exiting second washer (Site 4) . 51
24 Grit (inorganic solids) on turnip greens vs. accumulated
product at three sites in conventional system, trial 6,
Spring; unwashed product (Site 12), product exiting first
washer (Site 14), product exiting second washer (Site 15). 51
25 Total bacterial plate counts per gram of spinach at three
sampling points, trial 1, Spring; prototype system.
Before washing (Site 1), exiting the first washer
(Site 3), exiting the second washer (Site 4) 53
vii
-------
26 Total bacterial plate counts per gram of turnip greens at
three sampling points, trial 6, Spring; prototype system.
Before washing (Site 1), exiting the first washer (Site 3),
exiting the second washer (Site 4) 53
27 Total bacterial plate counts per gram of spinach at three
sampling points, trial 1, Spring; conventional system.
Before washing (Site 7), exiting the first washer (Site 9),
exiting the second washer (Site 10) 53
28 Total bacterial plate counts per gram of turnip greens of
three sampling points, trial 6, Spring; conventional system.
Before washing (Site 12), exiting the first washer (Site 14),
exiting the second washer (Site 15) 53
29 Bacterial populations and chlorine residual in wash water at
Site 1 of prototype, trial 2, Fall, when processing
collards 55
30 Bacterial populations and chlorine residual in wash water
at Site 4 of prototype, trial 2, Fall, when processing
collards 56
31 Total suspended solids vs. accumulated product input at all
six sampling sites, trial 4, Fall, when processing collards
with prototype system 58
32 Chemical oxygen demand vs. accumulated product at all six
sampling sites, trial 4, Spring, when processing turnip
greens with prototype system 58
33 Total suspended solids vs. accumulated product at all four
sampling sites, trial 1, Spring, spinach processed with
conventional washer 58
34 Chemical oxygen demand vs. accumulated product at all
four sites, trial 6, Spring, turnip greens processed
with conventional system 58
35 Five-day biochemical oxygen demand vs. color, trial 2,
Fall, when processing collards with prototype system .... 61
36 Five-day biochemical oxygen demand vs. color, trial 3,
Fall, when processing collards with prototype system .... 61
37 Five-day biochemical oxygen demand vs. color, trial 4,
Fall, when processing collards with prototype system .... 61
viii
-------
TABLES
Number Page
1 Summary of Information for Trials of Prototype Washer System
During the Fall Season of 1975 25
2 Summary of Information for Trials of Prototype and Conventional
Washing System During the Spring Season of 1976 26
3 Average Water Use Date for Prototype Leafy Vegetable Washing
System During Fall Trials, 1975 35
4 Average Water Use Data for Prototype Leafy Vegetable Washing
System During Spring Trials, 1976 36
5 Average Water Use Data for Conventional Leafy Vegetable
Washers During Spring Trials, 1976 39
6 Product Data for Prototype Leafy Vegetable Washing System
During Fall Trials, 1975 42
7 Product Data for Prototype Leafy Vegetable Washing System
During Spring Trials, 1976 43
8 Product Data for Conventional Leafy Vegetable Washers During
Spring Trials, 1976 44
9 Dry Weight of Grit from Various Units of the Prototype Leafy
Vegetable Washer at End of Each Trial 46
10 Accumulation of Floating Trash from Settling Tank Moving Belt
Screens for Prototype System During Fall and Spring Trials . . 49
11 Comparisons of Bacterial Population Densities (Total Plate Counts)
for Product Leaving to Product Entering a Two-Washer System
and for Water Leaving the Second Washer to Water Entering the
First During Greens - Washing Trials 54
12 Magnitude of Average Changes in Total Plate Counts From Beginning
to End of Trials at all Sampling Sites Recorded 57
13 Concentration of Phosdrin in Water of First Washer of Prototype
System , Spring Trials 60
ix
-------
Number Page
14 Waste Loads Discharged with Water from Prototype System During
Fall Trials, 1975 63
15 Waste Loads Discharged with Water from Prototype System During
Spring Trials, 1976 64
16 Waste Loads Discharged with Water from Conventional Washers
During Spring Trials, 1976 65
17 Waste Stream Characteristics From Prototype and Conventional
Systems 57
-------
ABBREVIATIONS
BOD — biochemical oxygen demand
BOD- — five-day biochemical oxygen demand
BOD2n — twenty-day biochemical oxygen demand
COD — chemical oxygen demand
TS — total solids
TSS — total suspended solids
VSS — volatile suspended solids
SS — suspended solids
02 — oxygen
metric ton — 1000 kilograms
PVC — polyvinylchloride
XI
-------
ACKNOWLEDGMENTS
The most generous cooperation of the personnel of Exmore Foods, Exmore,
Virginia is gratefully acknowledged. The open-handed willingness to allow
the use of plant space, utilities, and personnel time in the conduct of this
project was exemplary, indicating a far-sightedness that transcends immediate
gain. The results of this work will carry much additional weight by virtue
of the tests being performed in a practical, working environment.
Principals to be cited include Mr. Caspar Battaglia, President of Exmore
Foods, for his suggestion that his plant be used for the test site and his
continued interest throughout the project; Mr. Charles Floyd, Plant Manager,
for his cooperation in day-to-day arrangements; Mr. Stoakely Pearson, Plant
Engineer, for the skill, care, and energy exerted in getting the equipment
installed and making certain necessary modifications; and Mrs. Lucille Floyd,
Mr. Woodrow Brawley, and Mr. James Morrison for allowing the unrestrained
use of their laboratory. Many others should be cited, particularly the
foremen and workers on the processing lines. Their patience, humor, and
apparent pride in being associated with the project were a source of inspira-
tion, especially during trying moments.
Special recognition is extended to the graduate students associated
with the project, particularly Bill Robinson, Paige Geering, and Jim Coleman.
Several others also made significant contributions. The kind and amount
of work that they were subjected to and the inconveniences of the travel
imposed were uncommon compared to usual graduate studies. Their response
and enthusiasm were also uncommon — well above the ordinary requirements
implied by the receipt of stipends and degrees.
xii
-------
SECTION 1
INTRODUCTION
A 1971 estimate by the National Canners Associations indicated that the
1838 fruit and vegetable canning and freezing plants in the U. S. used 99
billion gallons of water and discharged 96 billion gallons of wastewater (14).
Approximately 626 million pounds of leafy greens and broccoli were processed
in 80 plants during that same year (6) (21) requiring an estimated 2.5 billion
gallons of process water. Even though greens processing represents only a
small percentage of the fruit and vegetable industry output, research in
areas related to it can have general applicability in many instances.
Two major concerns of leafy greens processors are water use management
and initial cleaning of freshly harvested product. Concerns in water
management, particularly those related to effluents, have assumed added
importance in recent years relative to the new emphases on environmental
protection. Major problems have arisen in handling effluents from the lack
of knowledge of waste stream characteristics. Design information on waste-
water parameters for treating combined flows from fruit and vegetable
processing is sketchy at best. Flow and concentrations of waste stream
constituents from unit operations within plants are even less available.
A limited amount of information is available on combined waste stream
loadings from leafy vegetable processing in reports by Mercer (12),
Ramseier (16), Frey (8) (9), the NCA (14) and SCS Engineers (18). Carter (4),
Bough (2), Frey and SCS Engineers have reported on certain unit operations.
The data available, however, are still inadequate for proper design of in-
plant or out-of-plant waste stream management. The parameters reported vary
from study to study and might include any of the following: BOD, COD, TS,
TSS, VSS, dissolved 02, pH, alkalinity, or bacterial counts. Methods of
reporting each parameter may also vary. For instance, COD may variously be
given in terms of miligrams/liter of wastewater, pounds per ton of product
processed or even pounds per 1000 cases of canned product. Other important
information, such as flow rates of product and water, is often omitted or
crudely estimated. Total water consumption has been reported to range from 3.2
to 5.4 gal/lb of greens processed. Estimates of total water consumption re-
quired for initial washing range from 68 to 88 percent. Obviously, the major
volume of the total wastewater comes from this source. While the inconsisten-
cies cited above do not necessarily invalidate reported results, they do limit
their usefulness and/or credibility.
Virtually no information is available on the relative effectiveness of
different devices used in the initial cleaning of various greens. Typical
-------
equipment used prior to blanching is described by Carter, Bough, Frey and
Lopez (11). This usually includes, in order, a dry tumbler for removal of
loose soil and small particles, hand inspection and picking belts, and from
one to four wet washers.
The present study was initiated to address the two major problems of
producing cleaner product in the initial processing of leafy greens and to
characterize the waste streams from these processes. An experimental, two-
washer, prototype system incorporating the principle of water recirculation
was constructed and tested during two harvesting seasons at a commercial,
frozen-vegetable processing plant. Design of the washers was based on
modifications of conventional washers developed by Frey to increase their
effectiveness. High recirculation rates within the system provided hydraulic
agitation to supplement the mechanical agitation. Input water was limited
to that required for makeup plus one small waste stream from the system.
Water and product quality and flow rates were monitored at several points in
the prototype during the fall of 1975 to determine washing effectiveness and
characterize internal and external water flows. A similar testing program,
conducted during the spring of 1976, included tests on conventional washers
for comparison purposes.
-------
SECTION 2
CONCLUSIONS
The experimental prototype leafy-greens washing system was more
effective, though not dramatically so, in removing grit and insects from
product than the conventional washers. It also showed a potential for
better control of bacteria counts on product prior to blanching. The final
rinse, with fresh chlorinated water, appeared to be quite effective for grit
removal and bacteria control. There was no apparent increase in grit or
insects on product as washing proceeded with recirculated water. A soap-
like foam accumulated on the water surfaces of the prototype that may have
had a significant effect on product cleaning.
Differences in water use between the two systems to obtain cleaning
was dramatic, the prototype using only about 1/5 the amount of water used by
the conventional washers. Waste water discharge from the prototype was
approximately 1/12 that of the conventional washers. The average amount of
water carried out on product from the prototype was 2.2 &/kg (0.26 gal/lb)
A fresh water input rate to the system of 3.5 A/kg (0.42 gal/lb) is a
recommended minimum.
The amount of each type of waste constituent (TSS, VSS, COD) discharged
with the water from the prototype system per unit of product processed was
less than that from the conventional washers though the concentrations were
higher. For example, the average discharges, from each system respectively,
while processing turnip greens were: TSS - 0.26 and 1.54, VSS - 0.04 and
0.26, COD - 0.16 and 2.31 kg per metric ton of product. Some of this differ-
ence, particularly for the non-volatile solids, was due to accumulations in
the washers and settling tanks of the prototype that could be disposed of
separately from the waste stream. VSS and COD production in the prototype
was less than in the conventional washers, probably due to lower osmotic
gradients between the recirculated water and the vegetables. Average con-
centrations in the discharge from the prototype and conventional washers,
respectively, while processing turnip greens were: TSS - 273 and 85, VSS -
37 and 14, COD - 135 and 128 mg/£ of waste stream.
Approximately 75 percent of the cleaning took place in the first washer-
settling tank sub-system of the prototype. Given steady inputs of product
and water similar to the average conditions of this study (1278 kg/hr and 72
Jl/min), the waste strength parameters in the washers and settling tanks will
stabilize at some maximum value after approximately 5 hours of operation.
This maximum value will be affected, of course, by the average "dirtiness"
of the vegetables.
-------
Waste production varies greatly between different varieties of vege-
tables and between different cuttings of the same vegetable. All para-
meters — organic, inorganic, insects and bacteria — are affected by age at
harvest, growing conditions, method of harvest, etc. Of the three varieties
tested, spinach consistently produced the most TSS per unit of product and
could be used as a model for design information for this waste parameter.
Other results, however, do not indicate that any of the three products
tested - collards, spinach or turnip greens — could be used as a general
model for VSS and COD emissions. Wash waters were generally neutral in all
trials indicating that pH would not be a problem in treatment of effluents.
The mechanical performance of the prototype washer was very satisfactory
though it could be improved as outlined in the recommendations section.
Of particular note are the product discharge conveyor belts and moving belt
screens for the recirculated water. The discharge belts were made of plastic
and appear to be a very effective and inexpensive substitute for stainless
steel. The screen belts, also of a monofilament plastic, provided a
relatively simple, inexpensive means of separating small leaf fragments and
even insects from the wash water.
Processors of frozen vegetables use large quantities of water to cool
blanched product prior to packaging. After the cooling water is separated
from the product it is usually used in the raw product washers. Assuming that
alternate means of economically cooling product (by chilled air, for example)
can be found, then freezers, as well as canners, of leafy-vegetables would
find considerable advantage in implementing low-water-use washing systems.
Recycling wash water in the initial processing of leafy vegetables is a
viable means of consolidating wastes, reducing the amount of effluent and
reducing the amount of total water required. Increased hydraulic agitation
of product by high internal flow rates in the system coupled with a final
rinse of controlled chlorine content can improve vegetable cleaning compared
to conventional washers. These findings are significant in terms of
environmental protection, resource conservation, and food quality. They
indicate that the final efforts needed to encourage implementation by the
food industry should be taken.
The intial cost of the prototype system developed in this study was
estimated at $16,000 compared to $12,000 for a conventional system of
equivalent capacity. Annual fixed costs of ownership were $2208 and $1656,
respectively,for a difference of $552 per year. Assuming a product mix of
3/4 spinach and 1/4 turnip greens, operating conditions similar to those in
this study, and using local labor and utility costs, the daily operating
cost for the prototype was $158 and for the conventional washers $251 an
advantage of $93 per day for the prototype. The difference in annual fixed
costs, in this example then, were recovered in approximately six days of
operation. If this is considered representative, then the economics of
owning and operating the two systems strongly favor the low-water-use
prototype washing system.
-------
SECTION 3
RECOMMENDATIONS
The prototype, leafy-vegetable washing system is effective in cleaning
leafy vegetables while using a minimum amount of water. No changes in its
functional design are considered to be necessary at this time. This does
not imply, however, that the effectiveness of the system could not be
improved by study of additional components or techniques in operation. The
relative effectiveness of the new system compared to similar conventional
washers does seem to warrant its adoption by the food processing industry
as soon as possible.
The present prototype has some limitations, unrelated to function,
that need improvement prior to considering it for commercial use over an
extended period. As now constructed, it requires too much space and is
too complicated. These problems, however, can easily be overcome by a
redesign that will not affect system performance and might possibly improve
it. For example, overflow water from each washer is now collected in a
sump and pumped over a moving-belt screen prior to discharge to a settling
tank. From the settling tank it is returned to the washer via a high pressure
spray system. The washer system could be greatly simplified by locating the
settling tanks and moving-belt screens underneath the washers where they could
receive the overflow by gravity. This would eliminate the sump pumps, reduce
the floor space required and would probably increase the effectiveness of the
settling tanks by equalizing the flow to them.
The redesign of the system should include construction of a second
prototype, some limited laboratory testing to verify certain operating
characteristics, and development of a complete set of plans and specifi-
cations. These plans could then be made available to interested food
processors and equipment manufacturers. The food processing industry as
a whole is very large and includes several giant corporations. Most food
processing plants, however, tend to operate in an autonomous fashion, draw-
ing little more than administrative support from their parent companies.
Research and development of needed machinery are pursued rather haphazardly,
usually on a "cut and try" basis. In order for a new system to achieve
maximum and speedy acceptance by this industry, information on it should
be presented in the most usable form. A processor given a complete set of
plans and specifications for an apparatus is more likely to build it in
his own shop or have it built than one who has to worry about design detail.
After the second prototype is built it should be installed in a commer-
cial processing plant, somewhere in the U. S., and used under normal operat-
ing conditions for an extended period of two to three years. This would
-------
provide a reference demonstration for other processors and allow for refine-
ments in design and operating technique.
Leafy vegetables go directly from the blancher to the cans in a canning
process. In frozen food plants, however, they must be cooled before being
packaged. This is usually done by fluming the product in large volumes of
fresh, cool water. The water from this cooling process, or a portion of it,
is then used in the washers after the cooled product is dewatered. Because the
cooling water is usually in excess of that required by the washers little
economic or environmental advantage would be gained by using low-water-use
washers in vegetable freezing plants without using alternate means of product
cooling. Some devices, such as air coolers, are available, but it appears
that further studies in this area of vegetable processing are warranted.
A comprehensive review of literature on both combined- and unit-
operation's effluents from fruit and vegetable processing should be
initiated before this literature becomes voluminous. This review should
be conducted with the object of accumulating known data in condensed form
and "normalizing" it to a standard form of presentation. A corollary
effort to this review would be the publication of guidelines for future
studies to indicate what data should be taken and how it should be
expressed. The review and guidelines should be developed with the view
of providing designers of processing equipment and waste treatment facil-
ities with the most useable data.
-------
SECTION 4
PROTOTYPE WASHER SYSTEM
WASHER DESIGN
Lopez (11) described the three most common types of leafy vegetable
washers as the 1) immersion, 2) rotary spray, and 3) spray belt. Of these
the immersion (sometimes called immersion drum, drum, paddle wheel or
dunker washer) is the most popular. Frey (8) (9), in response to industry
concerns for cleaner product, made several modifications to a conventional
immersion washer and demonstrated their effectiveness for removing insects
and grit from spinach. He also measured BOD, COD, TS, SS and VSS of the
waste stream. The water-use rate in this washer was approximately one gal/
Ib of product, well below the industry average. The low levels of the waste
strength parameters indicated the feasibility of developing a washer proto-
type system incorporating the principle of water recirculation.
General Design of Washing System
A full scale prototype of an immersion washing and water recirculating
system was constructed incorporating several of Frey's modifications to a
conventional washer. It consisted of two, modified, leafy-vegetable immer-
sion washers in series and their respective settling tanks and moving-belt
screens for cleaning the water in the recirculation process. Figure 1 shows
the arrangement of the system, as well as product- and water-flow patterns.
Fresh makeup water was introduced into settling tank number 2 during trials
in the fall of 1975. This arrangement was changed to apply the makeup water
as a final spray on the product leaving washer number 2 during the spring
trials of 1976. Excess water from settling tank number 2 overflowed into
settling tank number 1. This was the only hydraulic link between the two
washing units. Excess water in settling tank 1 flowed to waste. Figure
2 is an overhead view of the washing system as it was installed.
Description of a Washing Unit
Each washer and settling tank was constructed of 11 gage, type 304,
stainless steel sheets, with a 2 x 2 x 1/4 inch angle-iron frame around the
top. The washer was designed to wash approximately 4,000 pounds of product
per hour based on similar conventional washers at a local processing plant.
Each washer was 4 feet wide and 16 feet long with three "V" shaped sections
forming the bottom of the tank (Figure 3). This configuration aided in the
removal of grit when the tank was drained. The tank was three feet deep at
the deepest point, 1 foot-10 inches at the shallowest point, and held 688
gallons of water when filled to the working depth of 2 feet, 7 inches.
-------
1
WASI-
2
(£)
t
WASH
^^ ^
IER
•«-
ER
1
i
4
on -*•
GO
-j
t
i
j
FRESH
WATER
INPUT
SETTL
TA
i
\
SETTL
1
i
^.
®
.ING
NK 2
r i i
OVERFLOW FROM
SETTLING TANK 2
0) TO SETTLING TANK
.ING
'ANK 1
r
— PRODU
.. ... »i/A-rrr>
— ' •*• ««i ui\
Q PRODUCT S
O WATER SAh
O FLOW METI
DRAIN TO
WASTE
CT FLOW
FLOW
AMPLING SITES
rfPLING SITES
ERS
Figure 1: Diagram of washing system showing water and product flow
patterns, sampling sites and water flow meter locations.
8
-------
Figure 2. Overhead view of washing system adjacent to Exmore plant. Rotary sand
tumbler and conveyor into plant are at right and foreground.
-------
TOP VIEW
SIDE VIEW
A. Washer tank F. Ports for draining tank
B. Input conveyor G. Exit conveyor
C. Nozzle banks V. Water level
D. Paddle wheels with side drains and interior nozzles
E. Skimmers (later removed)
Figure 3: Diagram of Prototype Washer.
-------
Water was introduced from the settling tanks into the washer at several
locations. There were three banks of nozzles at the input end of the tank,
one located at the water level and two positioned above the incoming pro-
duct (Figure 3). Each bank consisted of four, brass, Flat-Jet No. 1/2
P35100 nozzles, manufactured by Spraying Systems Company. They were mounted
on 1-1/4-inch PVC pipe with split-eyelet connectors, Spraying Systems No.
8370A. These sprayers spread the incoming product, began the agitation
process to remove grit and trash, and propelled the product toward the first
agitation drum. The entire spraying system was designed for 200 gallons per
minute (gpm) at a pressure of 35 pounds per square inch (psi).
Three agitation drums, or paddle wheels, on each washer served to
agitate the product by alternately submerging and releasing it to remove
grit and trash. They were driven with No. 60 roller chain at 11 revolutions
per minute (rpm) by a 1-horsepower (hp), 3-phase electric motor coupled to
a Winsmith M300T right angle, 60:1, speed reducer. The drums were 1 foot
11-3/4 inches in diameter and were covered with 16 gage, 3/4-inch mesh,
flattened expanded stainless steel metal that allowed insects and leaf
fragments to float to the surface inside the drum while the product was
submerged (Figure 4). They each had four, 4-inch fins around their perimeter.
A stationary bank of three, flat-fan, brass Vee Jet No. H 1/2 U80100 nozzles,
manufactured by Spraying Systems Company was positioned inside each drum.
The nozzles were mounted with split-eyelet connectors on a 1-1/4 inch, PVC
pipe. This pipe was inserted through a hollow hub of the drum which located
the bank along the axis of the drum and thus allowed it to remain stationary
while the drum rotated. The nozzle bank was oriented so that the spray would
strike the drum covering at the water surface where the product was released.
This served to clean the drums during operation by preventing leaves from
becoming entangled in the expanded metal covering. In addition the spray
from these nozzles was another water input to the washer and an aid in the
agitation process. The drums propelled the product through the washer and on-
to an exit conveyor (Figures 3, 5). This conveyor was constructed of an open-
mesh belting made of plastic sections (manufactured by Intralox, Inc.). It
had flights every 24 inches and was driven with No. 60 roller chain by a
1/4-hp, single-phase, gear motor at a speed of 33 feet per minute (fpm).
The conveyor was inclined 30° from the horizontal.
The agitation drums had a spoked construction on one end (Figure 4) to
allow water and trash collected inside the drum to flow out through side
drains cut in the washer tank. The side drains were 4 1/2 inch diameter semi-
circles with their bottom edges located 5 inches below the top of the washer
(at the working depth of the water). Skimmers made of
3-inch diameter stainless steel tubing with lengthwise slots 4 feet long by
2 inches wide were installed directly behind each drum. They were positioned
to allow the product to flow beneath them before surfacing after it had been
submerged by the paddle wheels. Their purpose was to skim floating trash
from the water surface before it could recontaminate the product.
Water and trash from the skimmers and side drains (Figure 6) were
collected in a sump box, 2 feet by 2 feet by 1 foot 4 inches. A sump pump
(2-hp, 230 volt, single phase, Kenco No. 34N2 submersible), with a capacity
11
-------
N>
Figure 4. Paddle wheel showing expanded metal covering and spoked end construction,
Elevated nozzle banks are shown in foreground.
-------
Figure 5. Exit conveyor of washer number 2,
-------
Figure 6. View of washer side drains in operation.
14
-------
of 14,000 gallons per hour at 15 feet of head, was used to pump trash and
water from the sump box to a moving-belt screen that was mounted on top of
the settling tank. This pump was controlled by a Kenco Series 112-C12
Liquid Level Control.
A gate valve was used to regulate the flow from the sump pump through
a 3-inch, PVC pipe to the filter. The moving-belt screen was a conveyor (5
feet long by 1 foot wide) inclined at an angle of 16° from the horizontal
to prevent water from flowing off the exit end. The belt was made of
No. 410 Monofilament Polyester Screen manufactured by the Globe Albany
Company and had a permeability of 600 cubic feet per minute (cfm) per
square foot under a 1/4-inch head (Figure 7). The belt was chain-driven
at 47 fpm by a 1/4 hp electric gear motor. Trash was carried away on the
belt while the water flowed through it. A 3/4-inch galvanized pipe, which
had forty-eight 1/16-inch holes spaced 1/4-inch apart along its length,
was positioned under the exit end of the moving-belt screen. Compressed air
was directed through this pipe and against the belt to remove the trash. This
trash was collected in boxes placed at the end of the moving-belt screen
(Figure 8).
Water flowed through the moving-belt screen and into a settling tank
where grit could settle out. The tank was 8-feet long by 4-feet wide with a
4-foot-6 inch maximum depth and a 3-foot minimum depth. Figure 9 shows the
settling tank construction as well as the direction of water flow through it.
The baffles prevented floating material from getting to the pump. The tank
held approximately 700 gallons of water and had an overflow rate of 2.81 gpm/
ft^ based on an assumed particle size of 50 microns and a particle density of
2.65 g/cc [Metcalf and Eddy (13)].
An Aurora Model 344 centrifugal pump, with a 3-inch inlet and a 2 1/2-
inch outlet, was used to pump the water from the settling tank to the washer
spray nozzles. The pump capacity was 200 gpm against 35 psi. It was driven
by a 3-phase, 1800 rpm, 7-1/2 hp, electric motor with a V-belt drive.
A gate valve was used to regulate the flow from the settling tank
through a 2-1/2-inch PVC pipe to the washer tank. The flow was divided at
the washer and carried to the nozzle banks by 1-1/4-inch PVC pipe.
Motor starters for the 10 electric motors in the system were assembled
on a control panel, making it possible for one person to control the entire
washing system from one location.
WATER FLOW INSTRUMENTATION
Five meters were installed in the washing system to monitor fresh
water input, recirculation rates, and overflow rates (Figure 1). Meters
and 2 measured the recirculation rates within washing units 1 and 2,
respectively. The meters used were Badger Model MLFT, 3-inch totalizing
propeller meters with a normal operating range of 35 to 200 gpm.
15
-------
Figure 7. Moving 'Belt Screen in Operation
-------
Figure 8. Moving belt screen and trash collector. Compressed air hose for
removing trash from belt is shown in foreground.
-------
A. Water input
B. Water outlet to pump
C. Baffles
D. Direction of flow
E. Port for draining tank
Figure 9: Prototype settling tank
18
-------
Flow meter 3 measured the flow of fresh makeup water into the system.
A 1-1/2-inch Badger Model SC-ER totalizing, disc-type meter, with a normal
operating range between 5 and 80 gpm, was used to monitor this flow which
was regulated by a gate valve. The makeup water was piped directly into
the second settling tank from the meter during the test trials made in the
fall of 1975. During the spring trials of 1976, the makeup water was intro-
duced as a spray through five nozzles (Spraying Systems Co. Flat-Jet No. 1/2
P35100) onto the exit belt of washer 2 to provide a final product rinse.
This was the only modification made to the experimental prototype between
the two seasons.
The overflow from settling tank 2 to settling tank 1 was measured by
meter number 4. This meter was an 0.8-foot deep, Plexiglas, HS flume con-
structed according to specifications in the Field Manual for Research in
Agricultural Hydrology (10). A Friez FW-2, water stage recorder was used
to continuously monitor the depth of water in the flume (Figure 10). For
details of the construction and calibration of Plexiglas HS flumes, see
Robinson and Wright (17).
Meter number 5 measured the water flow from settling tank No. 1 to the
drain. An 0.8-foot, HS flume and Friez recorder were also used to monitor
this flow (Figure 11).
INSTALLATION AND MODIFICATIONS
The washing system was built in the Agricultural Engineering Department
Laboratory on the campus of Virginia Polytechnic Institute and State Univer-
sity (VPI&SU). It was determined to be operational and then disassembled and
transported to the Exmore Foods, Inc., plant in Exmore, Virginia. The
washing apparatus was reassembled adjacent to the Exmore plant for testing
at that site (Figure 2). Leafy greens were conveyed out of the plant to the
washers after having passed over a series of dry inspection belts. After
passing through the experimental washers, the product was carried back into
the plant and allowed to pass through the plant's conventional washers. A
reversible-feed belt was used to carry the product from the dry inspection
belts directly into the conventional washers when the experimental washers
were not in use.
Initial testing of the washer system with turnip greens revealed that
the skimmers did not function as anticipated. The product did not pass
under the skimmers, but collected on top of them. A skimmer had been used
successfully by Frey (8), but with a considerably lower product flow rate.
The skimmers were removed and their drains sealed off. The remaining side
drains, located at the ends of jthe paddle wheels, were not large enough to
carry off the flow introduced by the nozzle banks, so two modifications
were made to overcome this problem. Larger rectangular side-drains (4 inches
high by 7 inches wide) were cut at these locations with the bottom of each
drain 2 inches below the designed water surface level of the washer. In
addition, the bank of nozzles closest to the first drum was sealed off.
This latter modification was made in order to maintain high nozzle pressure
while reducing the water recirculation rate.
19
-------
Figure 10. HS flume meter number 4 with water level recorder
(Refer to Figure 1).
-------
Figure 11. HS flume meter number 5 with water level recorder
(Refer to Figure 1).
-------
SECTION 5
PROCEDURES
OVERVIEW
Construction of the prototype leafy green's washer was completed on
June 20, 1975, and attempts were made to purchase leafy vegetables in bulk
quantities for laboratory tests. Very dry weather in the local area, how-
ever, had shortened the spring harvest season considerably. Packers, who
would normally have excess leafy greens, were importing product from other
regions to meet their obligations. Hence, testing was postponed until the
fall season of 1975. Arrangements were made in the interim to test the
prototype on site at the Exmore Foods Plant in Exmore, Virginia.
Testing the washer in a commercial food plant had several advantages.
Tests were more realistic because the experimental equipment was subjected
to the same conditions under which conventional washers operate. The minimum
material needed for a reasonable test was estimated to be 25 tons. Arranging
for delivery of this amount of material to the laboratory without spoilage,
devising means to correctly meter it into the washers and disposing of it as
waste after testing would have been a formidable task. Pre-washing treatments
of product, such as dry tumbling and hand inspection, could also be performed
more easily in-plant than in the laboratory. A final advantage was that the
experimental equipment could be tested in comparison with conventional
washers. These comparison tests were subsequently arranged for the spring
season of 1976 under an extension of the original project.
Several difficulties were encountered as a result of working under
commercial conditions. Principal among these was the distance to the test
site, 350 miles. Each trip to collect data required a minimum of three
days, two for travel and one for tests. A large volume of samples was
taken during each trial, and it required special packing to avoid deteriora-
tion during transport from the plant back to the laboratory at Virginia
Tech. For each trial, a considerable number of small instruments and a
variety of glassware had to be transported to the plant and set up in a
temporary lab on the processing floor to supplement the company's laboratory
facilities. Finally, there was considerable difficulty in arranging travel
schedules to coincide with the plant's processing schedule, which was very
unpredictable. Decisions to process a certain vegetable, dependent on
weather and several other factors, were often made only a few hours in
advance of actual processing. Decisions to terminate processing were often
more precipitous, usually depending on some factor that affected quality.
22
-------
Additional difficulties, if they can be called that, included the fact
that the investigators had no control over the rate at which product was
processed, its initial condition before washing, or down time during trials.
Even input water flow rates fluctuated somewhat due to changes in operating
pressures in the plant's water system. Developing equipment and procedures
to control all of these variables in a laboratory experiment would, of
course, have added a degree of precision to the results, plus considerable
time and expense in obtaining them. This added precision, however, would
not have offset the insight gained from working under more realistic
conditions.
Prototype Installation, Plant Layout and Conventional Washers
The prototype washer system was installed adjacent to the Exmore Foods
Plant in Exmore, Virginia during the week of August 4, 1975 as described in
Section 4. Exmore Foods has two leafy vegetable processing lines, an east
line and a west line. The prototype was located so that it could operate in
series with the conventional washers of the west line or, when not in use,
could be bypassed (Figure 12).
The east and west conventional processing lines had two paddle wheel
washers each in series (Figure 12) followed by a combined paddle wheel
washer/pre-blancher, and then a blancher. The washers were three and one-
half feet wide, approximately eighteen feet long, and three feet deep at the
deepest point. There were four paddle wheels in each washer for propelling
the product through the washer. Product from the blancher was cooled and
transported in a cooling flume fed by fresh water. Dewatering and recircula-
tion of the cooling flume water provided all the input water for the washers
on the west line. The water input was at the head of each washer through a
perforated pipe. The dewatering of the cooling flume water on the east line
provided only a portion of the input water. The majority of this water was
fresh, piped into the bottom of the washers. All the overflow from the
washers was wasted to an open channel floor drain.
Overflow from the conventional washers was measured with HS flumes and
water stage recorders like those used in the prototype system. An attempt
was made to measure input water, which was under pressure, with propeller-
type totalizing water meters. These meters soon became inoperative because
some large, vegetable particles were being pumped from the dewaterers to the
washers. However, an accurate estimate of the input water, it was reasoned,
could be made by measuring the amounts of effluent water from each unit and
using the data in water carried off by the product available from the proto-
type trials.
Summary of Trials
Tables 1 and 2 summarize the trials that were made during the fall and
spring processing seasons. A total of 35,200 kilograms of product was
processed through the prototype in 27 hours of actual operation during the
fall and 31,500 kilograms in 24.4 hours during the spring. A total of
16,500 kilograms was processed in 11.7 hours through the east conventional
line and 21,200 kilograms in 15.3 hours through the west conventional line
23
-------
1
*
\
s
*
1
c
,s
SSS
(f
t
®
®
®
(7}
M.
\S\(>
)
V.
(-
01
. j:
t£
re
3
r-i
S
o
Convent!
I
— 1
sSS
Ei
T<
\v
1C
3
)
I
\s
1 of
Blai
ry :
_-
L\\\v
I
ac
I
\ <
.r
—1
sS
Jacl<
:her
:ast
spe
\\\V
a
\
:i
^\
ging
Line
ion
\\\\
W
•••
S11
Line
ielts
|
v\\\\\J
.-^ End of
Q§) Packaging Line
To Blancher
t
West Line
Outside Wall of Plant
Sane
Wagon
Tumbler
JL
Dry ".
ES
spe
ion
: ielt s
^
Reversible Feed '—
i\y&\\\\\\\\\\\\\\\\\V^
t
Figure 12: Schematic showing water and product sampling sites for
comparative study of new vs. conventional leafy greens
washing systems at Exmore Foods, Exmore, Va. Circled
numbers refer to water and/or product sampling sites.
24
-------
TABLE 1. 'SUMMARY OF INFORMATION FOR TRIALS OF PROTOTYPE WASHER SYSTEM
DURING THE FALL SEASON OF 1975
Trial
Date
Product
Total Operating
Time (hrs)
Total Fresh Weight
of Product Washed
(kg)
1
2
3
4
5
10/24/75
11/4/75
11/20/75
12/1/75
12/15/75
Collards
Collards
Collards
Collards
Spinach
3.80
5.78
6.68
6.68
4.05
**
4418
9975
8866
7945
3989
**
Fresh weight through washer system obtained from weight of product
packaged adjusted by determinations of relative moisture contents
of incoming and packaged product. These figures approximate raw
product entering first washer.
Product samples for moisture content analysis were damaged in
transport to laboratory. Estimate was made from data from other
collard trials.
25
-------
TABLE 2. SUMMARY OF INFORMATION FOR TRIALS OF PROTOTYPE AND CONVENTIONAL WASHING SYSTEM DURING THE
SPRING SEASON OF 1976
S3
ON
Trial
1
2
3
4
5
6
Date
4/22/76
5/12/76
5/21/76
6/4/76
6/10/76
6/11/76
Product
Spinach
Spinach
Spinach
Turnip
Greens
Turnip
Greens
Turnip
Greens
Total Operating
Prototype
6.83
-
-
5.50
6.17
5.92
Time (hrs)
Conventional
3.91 (E)
7.83 (E)
3.17 (W)
4.00 (W)
4.00 (W)
4.08 (W)
*
Total Fresh Weight
Prototype
8764
-
-
7999
6781
7974
of Product Washed (kg)
Conventional
4397
12132
5136
5525
5933
4658
No processing with prototype
Fresh weight through washer systems obtained from weight of product packaged adjusted by
determinations of relative moisture contents of incoming and packaged product. These figures
approximate weight of raw product entering first washer.
(E) = East conventional washing line (see Figure 12).
(W) = West conventional washing line (see Figure 12).
-------
during the spring. The amounts of product quoted here (and in the last
columns of Tables 1 and 2) are fresh (i.e. raw) weights as delivered into
the first washer from the dry inspection belts. Each trial consisted of a
complete or partial eight or nine hour shift. Down time and breaks were
substracted from total time.
The company packaged five different varieties of leafy greens during
the fall season, the largest volume of which was collards. Consequently,
travel schedules of the investigators coincided with collard processing
four out of five trials. The leaves of this variety tended to be large,
very mature, relatively clean, and easy to wash. Spinach was processed
during the fifth trial. It was not as clean as the collards, and the
leaves tended to be small and immature.
Initial plans for the spring season included dual trials with the
prototype on the west line running simultaneously with the conventional
east line. Unfortunately, there were only a few days early in the season
when both lines were processing the same product, and travel to the plant
could not be arranged at those times. As a best alternative, the prototype
system was tested during the day shift (8:00 a.m. to 5:00 p.m.), and data
were taken on one of the conventional lines the same night for the first
half of the night shift (6:30 p.m. to 11:00 p.m.). Clean-up and sample
preparation were usually complete by 12:30 a.m. Material processed during
a given day usually came from the same field, so comparisons between the
prototype and conventional lines could be made. Taking data on the con-
ventional washers for complete shifts was not necessary because the water
was not recirculated and, thence, effluent characteristics and product
quality parameters were not time dependent. Four trials were made in this
manner.
Data were taken on conventional lines alone during trials 2 and 3. The
west line, where the prototype was located, was not operated on the day of
trial 2. Plant operations were precipitously curtailed on the day of trial
3 by a deterioration in the quality of locally grown product. Only one
large truck load of spinach, purchased in a neighboring state, was processed.
Consequently, a decision was made to take data on the conventional washers
of the west line because data for spinach washing with the prototype were
already available. Quality of the spinach processed in the spring varied
considerablly—from prime (trial 3) to overly mature (trials 1 and 2) and
from very clean (trial 3) to very dirty because of sprinkler irrigation in
the field (trial 2). The variation in quality of turnip greens, although
great, was much less than spinach. Those in trial 4 were average in quality
and cleanliness, in trial 5 overly mature and clean, and in trial 6 good
quality and dirty.
SPECIFIC PROCEDURES
The following sections outline, in order, (1) water and product sampling
sites, (2) typical procedures that were followed for a trial during the fall
season, (3) modifications of those procedures for the spring trials and
(4) analytical procedures.
27
-------
Sampling Sites
Water and product sampling sites were selected so that the effect of
each major component of the system could be evaluated for each parameter
measured. Product samples were taken on the feed conveyor to the first
washer, the exit conveyor for the first washer and the exit conveyor for
the second washer in each case (product sampling sites 1, 2, 3, Figure 1
for the prototype; sites 7, 8, 10, for the east conventional line and 12,
13, 15 for the west conventional line, Figure 12). Packaged product samples
were taken at the end of the respective line in each trial (sites 11, 16,
Figure 12).
Water sampling sites for the prototype are depicted in Figure 1. Sites 1
and 3 were the spray nozzles at the head end of washers 1 and 2, respectively,
sites 2 and 4 the sump boxes that collected the total flow from each washer,
and sites 5 and 6 the input ends of the settling tanks. Hence, for example,
difference in samples between sites 1 and 2 measured the effect of washer 1,
between sites 2 and 6 the effect of the sump pump and filter belt, and between
sites 6 and 1 the effect of the settling tank. Similarly, water samples from
the conventional line were taken at entrance end of each washer and from the
overflow at the exit end of each washer. (Sites 7, 8, 9, 10, 12, 13, 14, 15,
Figure 12).
Typical Day - Fall Trials
The sequence of events for each trial, of course, varied.
outlines a typical work day during the trials.
The following
Time
Event
1) Night prior to trial
2) Morning of trial, before
processing
a) Laboratory set up on location.
b) Water quality instruments and flow
meters calibrated.
c) BOD dilution water prepared and aerated.
d) Prototype washers and settling tanks
rinsed and filled.
a) Washer and instruments turned on.
b) Flow rates set on meters 1, 2, and 3.
(See Figure 1).
c) Recorders started on meters 4 and 5.
d) Sodium sulfite prepared to neutralize
chlorine in BOD and bacteriological
samples.
28
-------
Time
3) Beginning of trial
4) After 15 minutes
operation
5) After each hour of
operating time (exclusive
of breaks and downtime).
Event
a) Start-up time recorded.
a) Grab samples of product taken for moist-
ure content, bacterial counts and insect
counts.
b) Grab samples of water taken for biochemical
oxygen demand (BOD), chemical oxygen demand
(COD), color, turbidity, chlorine residual,
total suspended solids (TSS), volatile
suspended solids (VSS), conductivity and
pH.
c) Grab samples of water collected in sterile
bottles for total plate counts and coli-
form counts (repeated at middle and end
of trial).
a) Same as 4, a and b.
b) Number of packages of product processed
recorded.
6) Between samplings
c) Flow meters 1, 2 and 3 checked with stop
watch.
a) Chlorine residual determined.
b) BOD's set up and placed in a low
temperature, Precision Model 815
incubator.
c) Conductivity, color, turbidity, and pH
determined.
d) Samples for TSS and VSS filtered, cruci-
bles and filters prepared for shipment
to Virginia Polytechnic Institute and
State University laboratory in Blacksburg,
Virginia.
e) Remaining water preserved with acid for
return to Blacksburg for COD determin-
ation.
f) Water samples for bacteriological
analysis plated and incubated at 35°C.
g) Product samples weighed and frozen in
plastic bags.
29
-------
Event
a) Trash collected and weighed.
b) Tanks drained.
c) Grit collected.
8) Day after trial a) Product samples packed in ice.
b) BOD bottles placed in cartons to
maintain approximate temperature of
20°C during transport.
c) Bateriological samples with sufficient
growth were counted. Others were packed
for transport.
d) Samples and equipment transported to
Blacksburg.
9) Week following trial (in a) Moisture contents, grit particle sizes,
V.P.I. & S.U. laboratory) COD's, TSS, VSS, BOD^s, BOD's, insect
counts and bacterial counts determined.
b) Preparations made for next trial.
Variations Between Spring and Fall Trials
Procedures for the spring trials were essentially the same as those for
the fall with the following exceptions:
(1) Product samples were taken for grit analysis in addition to samples
taken for moisture content, bacterial counts, and insect counts. Samples for
grit analysis were taken at two-hour intervals of operating time rather than
every hour. Product samples for grit analyses were hand washed and the wash
water filtered between sampling periods for later determination of suspended
solids.
(2) Water samples were taken for analyses of COD, TSS, VSS, chlorine
residual, pH, and pesticides both spring and fall, but BOD, color, turbidity,
and conductivity were not measured in the spring trials. Samples for
pesticide analysis were taken twice during each trial, mid-way and end, at
the outlet of the first washer of the prototype only (site 2, Figure 1).
During the fall trials water samples were taken at the input end of each
settling tank (sites 5 and 6, Figure 1) by dipping the container into the
surface of the water. Some rapidly settling solids may have been lost by
this technique. In the spring trials these samples were taken by holding
the container directly under the moving screen belt to catch the water before
it entered the settling tank.
30
-------
(3) Water and product sampling for the conventional washers were
essentially the same as those for the prototype except for sampling
associated with the settling tanks. An attempt was made to get a quanti-
tative analysis of grit accumulated in the conventional washers, but this
could not be accomplished without interfering with company personnel in
their clean-up procedures. One set of soil samples, however, was obtained
for particle size analysis.
Analytical Procedures
Following is a brief summary of the procedures used in analyzing samples
for each type df data taken during the investigation.
Water Flow Rates —
Totalizing water meters on the prototype were read at the beginning of
each trial and hourly thereafter to obtain flow rates as a function of time.
Depth of water flow through the HS flumes was continuously recorded on strip
chart recorders. Flow rates, as they varied with time, were later calculated
using these recordings and the calibration curve of flow vs. depth that had
been developed for the flumes (17).
Product Flow Rates
Product samples for moisture content determination were taken from input
conveyors, weighed, sealed in plastic bags and frozen for transport. Sample
packages of product from the end of the processing line were taken simul-
taneously and frozen for transport. At the lab, these samples were dried at
105°C for 24 hours in a forced convective oven, and moisture contents were
calculated on a wet basis. The relative moisture contents of input and
output product and the package counts of output product were then used to
calculate the rate of fresh product input to the washers. This was a
better measure of input product than total field weight because considerable
material was lost from the dry tumbler and the dry inspection belts ahead of
the washers and very little was removed by subsequent inspections between
the blanchers and packaging.
Grit Accumulation - -
The washers and settling tanks of the prototype were drained after
each trial and the volume of grit on the bottom of each tank was measured.
Some grit was inevitably lost while the tanks were draining. An attempt to
recover this loss was made by filtering the water as it flowed from the
tank. These filters were effective in trapping the larger sand particles,
but allowed some of the smaller silt and clay particles (those particles
with a diameter of less than 50 microns) to be lost. Grit loss in the
washer tanks was minimized by draining the tank through the port farthest
from the incoming product, where the grit accumulation was lowest. At the
end of trial 5 in the fall, a submersible sump pump was used to empty the
two settling tanks. This technique was used to empty the washers and
settling tanks in the spring trials.
31
-------
After the volume of grit had been determined, samples were taken from
each of the three bottom sections of the washers and from the settling
tanks. At the laboratory, they were dried in a forced convective oven at
105°C for 24 hours. A soil particle size analysis was then performed
using standard hydrometer methods (5) for particles in the range below
50 microns and standard sieve analysis (3) for larger particles.
Trash Accumulation - -
The trash collectors for the prototype were emptied at the end of each
trial and their contents weighed. Samples from each moving-belt screen were
taken, weighed, and frozen for transport to the laboratory. There, they were
dried by the same procedure as used for the product, and moisture contents
were calculated. These figures were compared with those for the incoming
product to determine the weight of trash collected when corrected to the
moisture content of the incoming product.
Water Sample Analyses - -
Following is a summary of methods used in analyzing water samples taken
during the trials.
Chlorine residual — The total chlorine residual was determined
amperometrically as described in Standard Methods for the Examination of
Water and Waste Water (Sec. 114B) (19) with a Fischer-Porter amperometric
titrator.
pH — The pH was determined as described in Standard Methods (Sec. 144A)
(19) with a Corning Model F pH Meter in the fall, with a Fisher Accumet Model
230 in the spring.
BOD,. — The BOD- was determined as described in Standard Methods (Sec.
219) (19). Sodium sulfite was added stoichiometrically to neutralize
the chlorine residual, which, if not treated, could kill the microorganisms
present in the sample. Experiments were conducted to determine the difference
between BOD,, of a seeded and unseeded sample. No difference was detected, so
the samples were not seeded.
Color — True color was determined by filtering a portion of water
sample through a Reeve Angel glass-fiber filter and determining the optical
density on a Klett-Summerson Photometer, using a #42 (blue) filter.
Turbidity — Turbidity was determined by measuring the optical density
of a small portion of water sample with a Klett-Summerson Photometer, using
a #42 (blue) filter.
Solids — Total suspended and volatile suspended solids were determined
according to Standard Methods (Sec. 1486 and 244D, respectively) (19).
COD — Water samples for COD were acid-fixed (ph ^2.0) for preservation
until analyzed. Determinations were made in the laboratory as described in
Standard Methods (Sec. 220) (19).
32
-------
Bacterial counts — The total plate count was determined according to
Standard Methods (Sec. 660) (19) except during trials 3, 4 and 5 during the
fall and trials 1, 2 and 3 during the spring when the streak-plate technique
was used instead of the pour-plate technique.
In addition to total plate counts, a non-specific coliform count was
made in the fall trials using desoxycholase agar, a selective medium for
coliforms. The pour-plate technique was used during trials 1 and 2 and
the streak-plate technique was used during trials 3, 4 and 5. All colonies
growing on the media were counted.
Pesticide Analyses — Pesticide analyses were made on water samples
taken from the first washer in the prototype during the spring trials. The
samples were examined only for organophosphorus pesticides, as they were the
only ones applied to the crops during the growing season. Phosdrin was the
principal one in use at the time of the study. Methods of analyses were
EPA-approved (Federal Register J38_, No. 85, Part II, Nov. 28, 1973), gas
chromatrographic analyses of solvent extracts of the samples. These were
conducted in the Department of Biochemistry Pesticide Analysis laboratory.
Product Sample Analyses —
Following is a summary of methods used in analyzing product samples
during the trials.
Bacterial counts — The method used for determining the total plate
count of the product was that used by technicians at Exmore Foods. Eleven
grams of product were placed in a bottle containing ninety milliliters of
sterile dilution water. From this bottle, a series of dilutions was made
and plated out on Total Plate Count agar. The test proceeded as described
in Standard Methods (Sec. 660) (19). Coliforms were determined on several
occasions (but not routinely) by plating aliquots of the water on desoxycho-
late agar.
Insect counts — Insect counts on product samples were made using the
gasoline extraction method described by Townsend et al. (20).
Grit on product — A simple hand washing test was devised to determine
the amount of grit on the product at each product sampling site. Duplicate,
1000-gram samples of product from each site were agitated 1-2 minutes by
hand in containers with 15 liters of water. Product was separated from the
water with a large-mesh screen and an aliquot of 500 ml. from each water
sample was filtered for solids determination.
33
-------
SECTION 6
RESULTS AND DISCUSSION
OPERATING PARAMETERS
The following is concerned with those quantities measured during the
trials that could be considered operating parameters, i.e., those things
that were either under the control of or required action by personnel
operating the systems. They include water and product flow rates for both
the prototype and conventional systems, and grit and trash accumulations
in the prototype system.
Water Flow Rates
Tables 3 and 4 summarize the water flow rates used in the prototype
during the fall and spring trials, respectively. Figures 13 and 14 are
examples of the water flows in the prototype for the fall and spring trials.
Meters 1 and 2 in each graph represent the recirculation rates for washers 1
and 2, respectively. Meter 3 is the input of fresh water to the system;
meter 4, the flow from settling tank 2 to settling tank 1; and meter 5, the
overflow to waste from settling tank 1 (refer to Figure 1). All tables and
graphs relate only to water flows during actual processing time. Meters 1, 2,
and 3 were totalizing meters and the graphs were plotted from readings taken
at timed intervals. Meters 4 and 5 were open channel HS flumes with water
level strip chart recorders. Data for the graphs was taken from the strip
charts at 15 minute intervals. Graphs for the spring trials were plotted by
hand (Figures 13, A-l through A-4). Graphs for the spring trials were taken
from computer plots (Figures 14, A-5 through A-12).
Recirculation rates in the washers varied somewhat but were generally
maintained near 404 £/min (107 gal/min). A recirculation rate of 530 Jl/min
(140 gal/min) in washer 1 was tried for a few hours in one trial (trial 5,
fall, Figure A-4) , but it was discovered that the control system on the
sump pump did not react fast enough to handle this flow.
Fresh water to the system was introduced into the second settling tank
during the fall trials (meter 3, Figure 1). No attempt was purposely made
to vary this flow, maintained as near the average of 66.8 A/min (17.7 gal/
mln) as fluctuations in the plant water pressure would permit. Fresh water
was introduced as a final rinse spray on the produce discharge belt of
washer 2 during the spring trials. In trials 4, 5, and 6 (Figures A-5, A-6,
A-7) this rate was increased with each trial [52.9, 71.3, 94.8 A/min (respec-
tively, 14.0, 18.8, 25.0 gal/min)]. During the fall trials, the input
water was turned off during periods of down time. This technique was tried
34
-------
CO
Ul
TABLE 3. AVERAGE WATER USE DATA FOR PROTOTYPE LEAFY VEGETABLE WASHING SYSTEM
DURING FALL TRIALS, 1975
Tr-f al
Nn
i
2
3
4
5
Date
10/24/75
11/4/75
11/20/75
12/1/75
12/15/75
Product
Cullards
Collards
Collards
Collards
Spinach
Averages
Recircu-
lation,
Washer 1
(«,/min)
398.6
437.1
366.9
362.8
483.1
410
(108 gal/
min)
Recircu-
1 at ion;
Washer 2
(X, /min):
333.5
392.2
377.5
368.7
479.2
390
(103 gal/
min)
Input to
System
(S,/min)
64.2
74.6
61.7
65.1
68.6
66.8
(17.7 gal/
min)
Output from
System
U/min)
38.2
20.0
18.9
25.8
25.5
25.7
(6.8 gal/
min)
Input
Water/
Product*
(A/kg)
3.31
2.59
2.79
3.28
4.18
3.23
(0.39 gal/
lb)
Water
Carried Out
On Product*
(A/kg)
1.34
1.90
1.94
1.98
2.63
1.96
(0.24 gal/
lb)
Fresh (raw) product entering first washer from dry insepction belts (Figure 12).
-------
TABLE 4. AVERAGE WATER USE DATA FOR PROTOTYPE LEAFY VEGETABLE WASHING SYSTEM
DURING SPRING TRIALS, 1976
u>
Trial
v. Date
NO.
1
4
5
6
4/22/76
6/4/76
6/10/76
6/11/76
Averages
Product
Spinach
Turnip
Greens
Turnip
Greens
Turnip
Greens
Recircu-
lation,
Washer 1
(S,/min)
376.8
405.4
360.7
443.1
397
(105 gal/
min)
Recircu-
lation,
Washer 2
(A /min)
448.3
453.8
377.1
397.7
419
(111 gal/
min)
Input to
System
(A/min)
92.0
52.9
71.3
94.8
77.8
(20.5 gal/
min)
Output from
System
(£/min)
29.5
3.9
24.1
37.7
23.8
(6.3 gal/
min)
Input
Water/
Product*
(A/kg)
4.30
2.18
3.99
4.22
3.68
(0.44 gal/
lb)
Water
Carried Out
On Product*
(A/kg)
2.92
2.02
2.64
2.54
2.53
(0.30 gal/
lb)
Fresh (raw) product entering first washer from dry inspection belts (Figure 12).
-------
420
400
380
360
1 340
1
I
320
300
80
6O
40
20
•METER I
•METER 2
-METERS
f 400
iH'kRATlKt. rim. (hrs.)
OPERATING TIME (HRS.)
Figure 13! Water flov races vs. operating tine, trial 1.' Fail,
1975. when processing collards with prototype
syatea. Refer to Figure 1 for meter locations.
Klgurp 14; Unter flow rotes vs. operating tin.c, tr
1976, when processing t.pinancli v-llh pro
Refer to Figure 1 for meter lor.it ton*.
500
^ 400
" 200
0 1234567"
OPERATING TIME (hrs.)
Figure 15. Wat*r overflow rates frost conventional washers vs.
operating tl»e, trial 1, spring, 1976, when
processing spinach on the east line.
37
-------
during trial 1 of the spring trials, but the amount of water carried off
by the product was such that there were periods of time after breaks when
the overflow to waste and the overflow between settling tanks was reduced
to zero (Figure 14). In trial 4 (Figure A-5) the input was left partially
open during breaks. Again, however, there were periods of no overflow to
waste. This was due to water carried off by the product plus the low rate
of fresh water input (52.9 £/min). On subsequent trials the fresh water was
left on during breaks which ultimately had the effect of reducing the con-
centration of the waste components used as measures of water quality.
Differences between the flow rates at meters 3 and 5 represent the water
being carried from the system on the product. Differences between meters
4 and 5 represent water carried from washer 1 into washer 2 by the product,
and differences between meters 3 and 4 represent additional wetting of the
product in washer 2. As expected, the amount of additional wetting of the
product in washer 2 was relatively slight in trials 1 and 2 during the fall
(Figures 13 and A-l). However, this trend was not observed in the remainder
of the fall and spring trials. This additional wetting may have been
influenced by a number of factors including the age and variety of product.
A particularly interesting phenomenon was the build-up of a soap-like foam
on the surface of the water in the settling tanks and washers. It resulted,
no doubt, from the surfactant action of organic matter leached from the greens.
Table 5 is a summary of the water flow data for the conventional washers,
and Figure 15 is an example graph (trial 1, spring). Graphs of the flows in
other trials are included in Figures A-8 through A-12, Appendix A.
The flow rate of water to the conventional washers was left entirely to
the judgement of the line foreman during each shift. More or less water was
used in each washer based on his judgement and experience. Thfe graphs show
considerable variations in flow that did not always appear to be related to
initial product quality or product flow rate. One mechanical influence on
the east line (trials 1 and 2, Figures 15 and A-8) was the inefficiency of
the cooling flume dewaterer, requiring that most of the input water to
these washers be fresh. Consequently, water use on this line tended to be
minimized, especially in washer 2. Conversely, the dewaterer on the west
line (trials 4 through 6, Figures A-9 through A-12) worked well, and all of
the input to these washers was recirculated product cooling water, used
unstintingly. In most, but not all, cases more water was used in the first
conventional washer in a line than in the second. It seems obvious that
some means of cooling blanched product other than with large quantities of
fresh water would be necessary in order for a company to take maximum
advantage of a water-conserving washing system.
Several other observations can be made from the water flow data. 1)
Though the number of trials for each product are few, there is an indication
that different varieties of greens tend to carry away different amounts of
water from a washing process. Average values, in order, are: collards -
1.79 A/kg (0.22 gal/lb), turnip greens - 2.40 £/kg (0.29 gal/lb), and
spinach - 2.78 A/kg (0.33 gal/lb). These figures, on a relative basis, are
consistent with expectations based on qualitative evaluations of leaf sur-
faces; i.e., collards have a waxy, smooth surface compared to spinach.
38
-------
TABLE 5. AVERAGE WATER USE DATA FOR CONVENTIONAL LEAFY
VEGETABLE WASHERS DURING SPRING TRIALS, 1976
Tria
J. Date
Product
and
Input tot
Two
° * Line Washers
1
2
3
4
5
6
4/22/76
5/12/76
5/21/76
6/4/76
6/10/76
6/11/76
Averages
Spinach(E)
Spinach(E)
Spinach (W)
Turnip
Greens (W)
Turnip
Greens (W)
Turnip
Greens (W)
(99
(£/min)
380.3
179.0
341.5
461.1
455.9
438.6
376
gal/min)
Output From
Two
Washers
(£/min)
328.3
107.2
266.3
405.9
396.6
393.0
316
(84 gal/min) (2
Input
Water/
Productf
(A/kg)
20.3
6.9
12.6
20.0
18.4
23.1
16.9
.02 gal/lb)
Water*
Carried Out
On Product+
CA/kg)
2.78
2.78
2.78
2.40
2.40
2.40
* Estimated from average values determined in prototype trials.
t Calculated from output data and estimated values of water carried out
on product.
(E) = East conventional line.
(W) = West conventional line.
+ Fresh (raw) product entering first washer from dry inspection belts
(Figure 12).
39
-------
2) The water input to two conventional washers averaged 5.2 times that of the
prototype system, the water output or waste stream 12.7 times that of the
prototype. 3) Average fresh water input for the conventional washers was
16.9 £/kg (2.02 gal/lb) vs. 3.43 A/kg (0.41 gal/lb) for the prototype, a
ratio of 5:1.
Product Flow Rates
Several different sizes of packages were used at the Exmore plant.
Retail packages for collards and turnip greens were 283 g nominal net weight
(10 oz) and those for spinach were 340 g (12 oz). In some cases diced
turnip roots were included with the greens (8 oz of greens, 2 oz of roots).
The weight of roots was accounted for when analyzing the moisture content of
samples, and only the greens processed are reported here. These packages
were counted with an electronic counter located immediately following the
packaging machine. Some product was packed in 6.8 kg (15 Ib) trays for
institutional packs and some on open trays in lots of 181 kg (400 Ib) for
bulk freezing and storage. The number of these units was recorded as they
were put into the freezer.
Figures 16 and 17 are examples of the instantaneous flow rate and cumu-
lative flow of fresh product into the washers versus operating time. Figures
A-13 through A-30 depict flows during other trials. Initially, it was
assumed that the flow of material through the system would be rather uniform.
Consequently, only the total product processed was recorded during trial 1
in the fall. During this trial it became obvious that input to the system
was very erratic. Inputs to the prototype system ranged from 456 to 2251
kg/hr (1105 to 4963 Ibs/hr) ; for the conventional washers 459 to 3330 kg/hr
(1012 to 7341 Ib/hr). These wide fluctuations in input rates undoubtedly
affected the quality of the product. An obvious means of improving washing
quality, and perhaps increasing the average processing rate, would be to
devise a means of metering the product more evenly into the washers.
Tables 6, 7, and 8 are summaries of the product data. The average rate
of fresh product input into the washers for all trials was 1324 kg/hr (2918
Ib/hr), and the average rate of output from processing was 1808 kg/hr (3985
Ib/hr), for an output/input ratio of 1.37. A limited number of tests during
the fall trails indicated that very little water was absorbed in the washing
process. It is assumed, therefore, that most of the water absorption took
place during blanching. The average input/output ratios for the three
varieties tested were: spinach - 1.54, collards - 1.34, and turnip greens -
1.25, indicating the relative abilities of each vegetable to absorb moisture
during processing. The range of absorption within each variety, however, was
considerable indicating that other factors—such as initial leaf condition
(turgid or wilted), age, size, etc.—would have effects. Particularly note-
worthy is the fact that variation in moisture content of the fresh product
(8.8%) was much greater than that of the packaged product (3.2%). Plant
records indicate that overall ratios for packaged product to product deliver-
ed from the field are approximately 0.75 for turnip greens, 0.90 for collards
and 0.95 for spinach. These ratios include wastage such as the losses on the
dry inspection belts and the gains due to water absorption. They vary con-
siderably both within seasons and from year to year.
40
-------
2500
2OOO
1500
1000
500
BREAK
LUNCH
BREAK
3 4
OPERATING TIME (MRS.)
Figure 16:
: Produce flow rate vs. operating time, trial 2, Fall. 1975, when prc
collards with the prototype systen. Fresli product into first washer.
ocessing
10000
9000
8000
700O
60OO
5000
4000
3000
2000
1000
01 234567
OPERATING TIME (MRS )
Figure 17: Arctinulatud product Input v*. npornUna. tint'. Trial J,
Ptllt 1975, when procrtmjng cutlardri with the prototype
iiyctrH. Freih product Into tlrnt wactier.
41
-------
TABLE 6. PRODUCT DATA FOR PROTOTYPE LEAFY VEGETABLE WASHING SYSTEM DURING FALL TRIALS, 1975
•is
to
Trial
No.
1
2
3
4
5
Date
10/24/75
11/4/75
11/20/75
12/1/75
12/15/75
Product
Collards
Co liar ds
Collards
Collards
Spinach
Avg. Fresh
Product"1"
Input
(kg/hr)
1163
1726
1327
1189
985
Avg. Moisture
Content of
Fresh Product
(% w.b.)*
88.6**
90.3
88.6
87.0
85.4
Avg. Moisture
Content of
Packaged
Product
(% w.b.)
91.3**
91.3**
91.2
91.4
90.3
Total
Packaged
Wt.+
(kg)
5891
12315
11514
11858
6137
Total
Fresh
Wt.
(kg)
4418
9975
8866
7945
3989
Ratio:
Packaged
Wt.
Fresh Wt.
1.33
1.24
1.30
1.49
1.54
Percent wet basis.
**
Estimate from other collard trials; samples lost in transit.
t Fresh (raw) product entering first washer from dry inspection belts, (Figure 12).
4 Divide packaged weight by 0.90 for collards and 0.95 for spinach to estimate field weight of
product delivered to plant.
-------
TABLE 7. PRODUCT DATA FOR PROTOTYPE LEAFY VEGETABLE WASHING SYSTEM DURING SPRING TRIALS. 1976
Trial
No.
1
4
5
6
Date
ft/22/76
6/4/76
6/10/76
6/11/76
Product
Spinach
Turnip
Greens
Turnip
Greens
Turnip
Greens
Avg. Fresh
Product f
Input
(kg/hr)
1283
1454
1099
1348
Avg. Moisture
Content of
Fresh Product
(% w.b.)*
84.6
90.3
90.9
91.2
Avg. Moisture
Content of
Packaged
Product.
(% w.b.)
90.0
93.1
**
91.9
**
92.1
Total
Packaged
Wt.+
(kg)
13426**
11189
**
7596
**
8916
Total
Fresh
Wt.
(kg)
8764
8000
6781
7974
Ratio:
Packaged
Wt.
Fresh Wt.
1.53
1.40
1.12
1.12
Percent wet basis.
**
Calculated from nominal package weight and number of packages.
t Fresh (raw) product entering first washer from dry inspection belts (Figure 12).
* Divide packaged weight by 0.95 for spinach and 0.75 for turnip greens to estimate field weight
of product delivered to plant.
-------
TABLE 8. PRODUCT DATA FOR CONVENTIONAL LEAFY VEGETABLE WASHERS DURING SPRING TRIALS. 1976
Trial
No.
1
2
3
4
5
6
Date
4/22/76
5/12/76
5/21/76
6/4/76
6/10/76
6/11/76
Product
Spinach
Spinach
Spinach
Turnip
Greens
Turnip
Greens
Turnip
Greens
Avg. Fresh
Product
Input t
(kg/hr)
1122
1549
1622
1381
1483
1141
Avg. Moisture
Content of
Fresh Product
(% w.b.)
82.4
86.1
90.8
89.4
90.8
89.7
Avg. Moisture
Content of
Packaged
Product
(% w.b.)*
90.1
91.1
93.0
93.2
**
92.3
**
92.7
Total
Packaged
Wt.*
(kg)
7802
18754
6713
8438
**
7064
**
6525
Total
Fresh
Wt.
(kg)
4397
12132
5136
5525
5933
4658
Ratio :
Packaged
Wt.
Fresh Wt.
1.78
1.55
1.31
1.53
1.19
1.12
Percent wet basis.
**
Calculated from nominal package weight and number of packages.
t Fresh (raw) product entering first washer from dry inspection belts (Figure 12).
* Divide packaged weight by 0.95 for spinach and 0.75 for turnip greens to estimate field weight
of product delivered to plant.
-------
Grit Accumulation
Removing accumulated grit from the washers is one of the clean-up tasks
required at the end of a shift, or sooner if necessary. Mechanical or
hydraulic means for continuous removal could have heen incorporated into the
experimental system at considerable expense. An expedient, but workable,
alternative used in the operation of the conventional washers was to open
the first drain valve on the first washer for a few seconds as necessary to
"flush out" excess grit, because most grit tends to settle out immediately
beneath the fresh product input. This technique was not often required,
although it did increase the effort needed in clean-up when used.
The amount of grit on incoming product varies greatly depending on the
particular type green being processed. Spinach is usually the "dirtiest"
because the convolutions in the leaf surfaces tend to trap soil particles
and because it grows close to the ground. Collards, on the other hand, have
smooth, waxy surfaces and grow erect. Turnips greens are intermediate in
these characteristics. Other factors include soil splashing from recent
rains or sprinkler irrigation, age of the leaves (older leaves tend to be
larger, smoother and cleaner), and the cutting (first cuttings are made
closer to the ground than subsequent ones).
Table 9 summarizes the measurements of accumulated grit in the trials of
the prototype washer. The collards processed during the fall trials 1
through 4 were very clean, leaving very little grit in the system. During
these trials, several techniques were tried to capture or collect the grit
as the water flowed out of the drains. In trial 5 (spinach) a measurable
amount of grit accumulated in the system, and it was scooped out of each
tank after the tank had been drained over the top with a sump pump. This
technique was used in the spring trials with the prototype and required
approximately an hour's work for each clean-up plus considerable agility on
the part of the clean-up crew.
The majority of the grit collection took place in the first washer (38%
avg.) and in the first settling tank (43% avg.). Only 4 percent was collected
in the second washer and 15 percent in the second settling tank. Although
the figures for the maximum amount in each trial vary between washer 1 and
settling tank 1, the amount accumulated in washer 2 was always the lowest
for all four units. In only one case (trial 6, spring) did settling tank 2
collect more grit than settling tank 1. These figures strongly indicate
that the majority of grit removal took place in the first washer sub-system.
The washers each had three drains, located at the apex of the V-shaped
bottom sections (Figure 3). Approximately 60 percent of the grit collected
in each of the washers settled in the first V-section, 26 percent in the
second and 13 percent in the third. Figures 18 and 19 show the summation
percentages of the particle size analyses for grit from the various units of
the prototype system for trial 5 in the fall. Figures A-31 through A-34,
Appendix A, depict these results for the spring trials. These analyses
indicate: (1) that most of the larger particles (>100 y) settle out in the
washers, (2) that the settling tanks were removing some particulate matter
smaller than the design diameter of 50 y, and (3) there was little or no
45
-------
TABLE 9. DRY WEIGHT OF GRIT FROM VARIOUS UNITS OF THE PROTOTYPE LEAFY VEGETABLE
WASHER AT END OF EACH TRIAL*
Trial
No.
5*
.**
4
5
6
Date
12/15/75
4/22/76
6/4/76
6/10/76
Wll/76
Product
Spinach
Spinach
Turnip
Greens
Turnip
Greens
Turnip
Greens
Washer
1
(kg)
8.5
16.9
10.4
4.4
17.9
Washer
2
(kg)
0.4
3.9
0.7
0.3
1.3
Settling
Tank 1
(kg)
38.3
19.1
4.7
2.4
1.1
Settling
Tank 2
(kg)
12.7
6.9
1.3
0.6
2.4
System
Total
(kg)
59.9
46.8
17.1
7.7
22.7
Total/
Product t
(g/kg)
15.0
5.3
2.1
1.1
2.9
Grit settling in washing system during fall trials 1-4 (collards) was very slight.
**
Spring trials.
± Does not include grit discharge from system during each trial.
t Fresh (raw) product entering first washer from dry inspection belts (Figure 12).
-------
O WASHER 1 SECTION 1
MASHER 1 SECTION 2
» SKTTLISC TASK I
a WASHER NO. 2 SECTION NO. 1
0 SETTLING TANK SO. 2
PARTICLE SUE IK MICRONS
PARTICLE SIZES IS MICRONS
Figure IB: Sunnaclon percentages vs. particle size for grit W'V "<••) accuoulatcd In the prototype Flgur. 19. Suo,uclo
•ysccn sub-unit 1; Trial 5, Fall, 1975, uhen processing spinach.
tide sUe for grit (dry ut.) accumulated in the ?r--u-:yp«
sub-unit J, Trial 5. FjLL. 1975, whun procesidlng spinach.
20 60 611 rill IUO JIXI
PARTICLE SIZi: IN Mli:RiVo
Figure 20: Summation purr.cnt.iK<"; vs. particlti slzo for grit (dr; wi.) i,aiii|)li^ lak^u (run vtnwcnti.mil
Uasher 1, Eaut l.lm>, TrUI 2, Spring. l*)7l> wlir>n prorc^slnR splnjdi.
-------
variation in the particle size distribution of grit from spinach, collards or
turnip greens. These analyses should be useful for designers developing
similar equipment for product grown on sandy and sandy-loam soils. For
other soils, particle-size analyses would have to be developed.
The effectiveness of grit removal in the settling tanks could probably
be improved by using a lower recirculation rate. Necessary hydraulic agita-
tion could be provided by increasing the pressure on the spray nozzles in
the washers. The surges in flow to the settling tanks produced by the sump
pumps probably reduced effectiveness of the tanks. This problem could be
eliminated by locating the settling tanks beneath the washers so that they
would receive the overflow by gravity.
The amount of time required to remove accumulated grit, and the con-
figuration of the conventional washers made it impossible to measure the grit
collected in them during the trials. Only in trial 2 (spring) was the amount
of grit on incoming product significantly different from that processed
during the prototype runs. Accumulated grit had to be shoveled out of the
first conventional washer during the lunch break—approximately 1700 kg.
Visual observation of the conventional washers in the other trials indicated
that the amount collected was of the same order of magnitude as that collected
in the prototype washers. Again, most of this accumulation was in the front
section of the first washer. Only limited amounts accumulated in the second
washer. Samples for particle size analysis were taken during trial 2 (Figure
20). Because the accumulated soil had reached a level within a few inches
of the water level in the washer where smaller particles could be deposited
it is assumed that these results are representative of the total soil brought
in on the product. If true then, 86 percent of the grit to be removed from
the product was in the size range above 100 microns.
Trash Accumulation
Two operational characteristics can be derived from the trash collection
data for the prototype system. They are: (1) the amount of waste product
that would have to be disposed of and (2) the amount expected to be lost
during washing. Table 10 summarizes these data and indicates that neither
of the above would be a major concern. The data for equivalent weight of
fresh product lost to trash was determined by adjusting the wet weight.of
the trash to the average moisture content of the incoming product during
each trial.
The trash generated by the prototype consisted primarily of leaf
particles. This type of material from the conventional washer flowed
directly into the waste stream and had to be removed by vibrating screens
before the plant effluent was released to treatment facilities.
PRODUCT QUALITY PARAMETERS
The product quality parameters measured were insect counts, bacterial
contamination and grit on the leaves. Of these, only insect and bacterial
48
-------
VO
TABLE 10. ACCUMULATION OF FLOATING TRASH FROM SETTLING TANK MOVING BELT SCREENS FOR
PROTOTYPE SYSTEM DURING FALL AND SPRING TRIALS
Trial
No.
2
3
4
5
1
4
5
6
Date
11/4/75
11/20/75
12/1/75
12/15/75
4/22/76
6/4/76
6/10/76
6/11/76
Product
Collards
Collar ds
Collards
Spinach.
Spinach
Turnip
Greens
Turnip
Greens
Turnip
Greens
Moving
Belt
Screen
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
Total Wet
Wt.
(kg)
14.1
14.1
18.5
24.2
14.2**
14.0**
25.4
31.3
29.9
43.1
5.1
13.6
3.4
10.9
10.9
18.6
Wet Trash/
Productt
(g/kg)
0.2
4.8
3.6
14.2
8.3
2.3
2.1
3.7
Equivalent
Wt. of Fresh
Product
(kg)
*
11.9
15.2
8.0
7.9
16.8
16.4
24.7
30.0
4.9
11.6
4.3
9.1
11.3
13.4
Equivalent
Product
(g/kg)
*
3.1
2.0
8.3
6.2
2.1
2.0
3.1
**
Moisture content samples damaged in transit.
k
Trash collectors damaged by windstorm after 1.76 hours operation in a.m. Reinstalled in p.m.
Trash collection for intervening time (1.70 hours) estimated from collection during rest of day.
t Fresh (raw) product entering first washer from dry inspection belts (Figure 12).
-------
counts were made during the fall trials. The hand washing test for grit
analysis was not devised until the start of the spring trials.
Insect Counts
Insect infestation of product was extremely low during the fall trials
and not very severe during the spring trials. Though fortunate for Exmore
Foods, this situation did not allow for a very rigorous test of the proto-
type for removing insects. Insect counts for trial 5 (fall) and trials
1-6 (spring) are included in Tables B-l through B-7 in Appendix B. A total
of only 10 insects were found on all the product samples taken during trials
1 through 4 in the fall with no more than 2 in any one sample.
Significant insect populations on the product appeared only during
trial 2 (spring) when the prototype was not in operation. In this trial
the two conventional washers removed 63% of the insects (aphids) from the
spinach, and for all trials averaged 62%. Whether or not this could be
considered representative is not known.
Several bits of evidence indicate that the prototype was effective in
removing insects. First, the data show that it was consistent in lowering
insect counts as the product was washed. Second, there was no evidence of
build-up of insects or insect fragments on the product as time proceeded.
This is significant in view of the fact that the wash water was being
recirculated. Finally, counts on the trash collected from the washers
during the fall averaged 124 whole insects and 81 fragments per 100 grams
even though incoming product averaged only 4 whole and 1 fragment per 100
grams of sample taken. Over all trials the prototype removed 70% of all
insects.
Grit on Product
Figures 21 through 24 depict some of the results of the hand washing
tests of product samples for grit analysis in the spring trials. Complete
data are included in Tables B-8 through B-13, Appendix B. The amount of
grit on entering product varied greatly. The amount on product leaving the
first washer varied less than initial readings and that leaving the second
washer even less. In all four trials where both the prototype and conventional
systems were operated the prototype removed a greater percentage of the grit -
averages of 73 percent and 69 percent respectively.
Spinach harbored more grit than turnip greens. In trial 1 the prototype
removed 80% of the grit and the conventional washers 73%. The conventional
washers averaged 70% removal in trials 1, 2 and 3 (spinach). The maximum
amount of grit observed on incoming spinach was 22,000 mg per lOOOg product,
10 times greater than the maximum observation for turnip greens, 2070 mg per
lOOOg of product. In trials 4, 5, and 6 (turnip greens) the prototype aver-
aged 64 percent removal and the conventional washers 59 percent. Removal
percentages for turnip greens appeared to be lower in general than those for
spinach. This probably relates more to the amount of grit on product than
any other factor in these trials. Given several varieties of greens of "equal
dirtiness" spinach would undoubtedly be the most difficult to clean.
50
-------
6 O
8 9
or
x 3
ACCUMULATED PRODUCT INPUT, KG X IO
0 I 2 3 4 5
ACCUMULATED PRODUCT INPUT, KG X IQ3
Figure 21: Grit (Inorganic solids) on spinach vs. accumulated
product at three sites in prototype svstcn, Trial 1,
Spring; unwashed product (Site 1), product exiting
first washer (Site 3), produce exiting KCCond washer
(Site «).
Kigure 22: i:rlt (innrgani
.!-) .-ii •.pln.v'i v... ji ior
(Site 10).
160
O 13 5
o
o
§ 9-0
4-5
24
I
o
o
o
(8
I. 2 345678
ACCUMULATED PRODUCT INPUT, KG X 10
0 1 2 3 4 35
ACCUMULATED PRODUCT INPUT. KGXIO
Figure 23; Crft (Inorganic solids) on turnip green* vs.
accumulated product at three slccs in prototype
svitten. Trial 6, Spring; unwashed product (Rite I),
product exiting first vanher (Site 3), product
exiting second washer (Site 4).
Figure 24: (-rlt (liturganlr soli
accunulatcd product
system, Trljl d. Spr
pruduct cxltitiH (Irs
exiting second was lie
•0 on turnip ftrecti5 vs.
three sites In conventional
nit; unwashed product (Site 12),
v.ixhcr (Site 14), product
(Site 15).
51
-------
The greater percentage removal of grit by the prototype over the conven-
tional system may be attributed to the increased hydraulic agitation and
the final fresh water rinse. However there was a decrease in the percent
of grit removal with time in three of four trials, probably due to the use
of recirculated water. Grit removal in the conventional washers varied
randomly with time.
Bacteria Counts
Bacterial counts for both product and water samples are not as complete
as originally planned for a variety of reasons — including considerable
variation in the amount of bacteria on input product, problems in transporting
samples under controlled conditions and, in general, the rather primitive
laboratory facilities on site. The data, however, are sufficient for certain
inferences.
Bacteria counts on product samples are included in Tables B-14 through
B-23, Appendix B. Table 11 is a summary of the principal effects on product
and water bacteria. Over all trials, the prototype and the conventional
systems each reduced the bacterial counts on product 74 percent of the time.
Of special significance, however, are the prototype trials in the spring —
1, 4, 5, 6. Unfortunately, a spreader-type organism present in the water
overgrew the plates in trial 4, obscuring the colonies. In the other three
trials, however, there was a very consistent and obvious lowering of bacterial
counts (100% of the time) as the product passed through the system. Undoubt-
edly the addition of fresh chlorinated water as a final rinse on the product
in these trials had an influence here.
Factors other than chlorine in the wash water appear to be operating
in the reduction of bacteria on product. It may be that something from the
product itself, which accumulates in the water, has a bacteriacidal action.
Again, the results of the spring trials with the prototype showed a consistent
reduction in counts in the first washer which received no fresh water (Figures
25 and 26). Counts for the conventional washers did not exhibit this
consistency (Figures 27 and 28). Two other observations can be made
from the data on bacteria counts. 1) There does not seem to be a clear
relationship, if any at all, between counts in the wash water (data presented
below) and on the product. 2) Though not obvious, it appears that higher
initial bacteria counts may be expected on spinach than on collards or turnip
greens. There are many uncontrolled influences here, however.
WATER QUALITY PARAMETERS
A total of 10 different water quality parameters were measured during the
trials. In the fall these included bacteria counts (including total coliforms),
chlorine residual, BOD,, BOD , COD, TSS, VSS, turbidity, color and pH. In
the spring bacteria counts, chlorine residual, COD, TSS, VSS and pesticide
analyses were made.
52
-------
250
20O
1
s
CE 150
£
"o
x too
w
UJ
Z
8 50
o
.
1
1
«
1
^
1
!
«
\
SITE 134. . 1 3 4. . 1 3 4,
TIME .25HOUR 4HOUR 7 HOUR
UJ
I
o
400
30O
200
IOO
o
•
1
!
SITE .1
i
3
1
4
TIME -25 HOUR
j
1
1
,\
\
y
v
\
\
I '
V, TOi rm
.1 34. ,1 3 4.
4 HOUR 65 HOUR
Figure 25: Total bacterial plate count* per gran of spinach at
three sampling points. Trial 1, Spring; prototype system.
Before washing (Site 1), exiting the first washer
(Sice 3), exiting the second washer (Sice 4).
Figure 26: Total bacterial plate counts per Rr.im of turnip (trro
at three sampling points. Trial A. Spring; prat»tvpe
system. Before washIns (Site 1), exiting tlu- first
washer (Site 3). exiting the second washer (Site A).
z
o
?00
ISO
100
50
n
•
I
I
^
\\
•
SITE .7 9 10. 7 9 10.
TIME -25 HOUR 4 HOUR
30O
S
a:
o
K
"• 200
9
x
V)
1J
z
3 100
SITE .1?
TIME .25 HOUR
I HOUR
Figure 27: Total bacterial plate councft per gran of spinach
ac three sampling points. Trial 1, Spring; convenclonal
system. Before washing (Site 7), exiting Che first
washer (Sice 9), exiting Che second washer (Site 10).
Figure 28: Total bacterial plate counts per gram of turnip greens
of three sampling points, Trial 6, Spring; conventional
RVRtcm. Before washing (Site 12), exiting the first
w.isher (Site 14), exiting Che second washer (Sice 15).
53
-------
TABLE 11. COMPARISONS OF BACTERIAL POPULATION DENSITIES (TOTAL PLATE COUNTS)
FOR PRODUCT LEAVING TO PRODUCT ENTERING A TWO-WASHER SYSTEM AND
FOR WATER LEAVING THE SECOND WASHER TO WATER ENTERING THE FIRST
DURING GREENS - WASHING TRIALS.
Trial
Product
Water
A* B**
Product
A* B**
***
Total Chlorine, mg/L
Beginning
End Average
Fall trials, prototype washer
1
2
3
4
5
Spring
1
5
6
Spring
I
2
3
4
5
6
Collards
Collarda
Collards
Collards
Spinach
3/3
3/3
3/3
2/2
2/3
0.15
0.02
0.23
0.09
0.98
-
1/4
3/4
6/8
3/5
2
0
0
0
-
.17
.46
.64
.50
2.
2.
0.
1.
1.
16
58
4
16
30
0.65
0.0
2.03
2.30
1.50
1
1
1
2
1
.10
.39
.35
.06
.52
trials, prototype washer
Spinach
Turnip
Greens
Turnip
Greens
3/3
2/3
2/2
0.11
0.73
0.28
3/3
3/3
3/3
0
0
0
.72
.33
.07
1.
1.
1.
30
79
32
0.57
0.68
0.62
0
1
0
.61
.02
.91
trials, conventional washers
Spinach
Spinach
Spinach
Turnip
Greens
Turnip
Greens
Turnip
Greens
2/3
-
1/3
1/2
1/3
1/3
2.57
-
1.49
1.15
i.22
0.92
2/2
2/3
3/3
4/5
3/5
3/5
0
0
0
6
0
0
.59
.04
.03
.26
.38
.85
1.
0.
0.
0.
1.
0.
23
28
23
55
56
54
1.15
0
0
0
0
0
1
0
0
0
0
0
.07
.47
.08
.18
.52
.18
*
A =
ratio of number of observations during a trial that the bacterial count
was lowered (input of first washer to output of second washer) to total
number of observations.
B " ratio of average bacteria counts during trial, output of second washer
to input of first.
*** • Residuals measured amperometrically and represent concentrations at the
back of the second washer at the beginning and end of each trial, and
the average for the entire trial. (3 to 8 observations per trial).
54
-------
Bacteria Counts and Chlorine Residual
Total plate count data are recorded in tables C-l through C-10, Appendix
C. These counts ranged over 5 orders of magnitude during the fall trials and
4 orders of magnitude in the spring. The relative counts for water where the
product entered compared to those where it was discharged in the prototype
system exhibited some consistency as shown in the first two columns of Table
11. The bacteria count was lower for the output of the second washer compared
to the input of the first washer in 20 of 22 observations. The opposite was
true in the conventional washers where the output counts were lower only for
6 of 14 observations. These last results seem to indicate that the last water
the product is exposed to does not necessarily have a direct effect on the
product bacteri^. counts which were consistently lowered by the conventional
washers.
The response to chlorine was not always consistent either as indicated
in Figures 29 and 30. Total chlorine was measured and recorded each time a
water sample was taken. Chlorine input to the fresh water in the plant was
manually regulated and varied considerably. A high value of 3.8 mg/Jl was re-
corded in the fall trials and a high of 1.79 mg/£ in the spring. Concentra-
tions usually but not always decreased toward the end of the trials (Table 11),
often falling to zero. This correlates with the fact that 73 percent of the
counts on water samples taken at the end of the trials (all sites) were
greater than those taken at the beginning. Fewer (54 percent) of the product
samples at those times showed higher counts, and the magnitude of these
changes was not as great as for the water samples.
The data in Table 12 re-enforce the lack of correlation between bacteria
in water and on product. There appeared to be a consistent increase in
bacteria counts in the water in both systems as time proceeded. The build-up
in the conventional system may indicate a trend to higher levels because
these trials were usually shorter than those with the prototype. The
differences in counts on the product for the fall and spring prototype trials
may be attributed to the generally higher levels of chlorine in the water
during the fall. Warmer weather in the spring may have also been a factor.
The differences between the build-up of counts on product for the prototype
and conventional washers in the spring may merely be the influence of exposing
the product to more chlorine treated water in the conventional washers.
In summary the influences on bacteria counts, both product and water, are not
clear. A system similar to the prototype using a final product rinse with
closely controlled chlorine content would appear, however, to have an
advantage over conventional washers in bacteria control.
55
-------
Ul
8-7 4
O 3
O PLATE COUNT
A CHLORINE RESIDUAL
6.0
5.5
5.0
45
4.0 :
I
3.5
3.0 '
IB i
20
1.5
1.0
0.50
"0246 ~§ JO 12 J4 i6~
ACCUMULATED PRODUCT INPUT,KG.XIO3
29: BJCU-rinl populatlnnn /ind chlorine residual In wash wnter
at Site 1 of prototype, irlul 2. Fall, when processing
1.0
"0 2 4 6 8 10 12 14 16
ACCUMULATED PRODUCT INPUT, KG XIO3
Figure 30: B4H.lerial populations and chlorine reaidutl in waah
walcr at Sice 4 of prototype, trial 2, Fall, when
-------
TABLE 12. MAGNITUDE OF AVERAGE CHANGES IN TOTAL PLATE COUNTS FROM BEGINNING
TO END OF TRIALS AT ALL SAMPLING SITES RECORDED.
Trials
Prototype (Fall)
Prototype (Spring)
Conventional
Water*
+169.0
+ 78.3
+356.0
Product**
-110.3
+ 41.7
- 4.6
* 3
Colonies X 10 per milliliter
** 3
Colonies X 10 per gram
Tables C-ll through C-15, Appendix C record data from the fall trials
on total coliforms in the wash water. These organisms indicate the presence
of fecal material on the incoming product. Their effect on final product,
however, is not known.
TSS. VSS. COD and BOD
Figures 31 and 32 are two examples of water quality parameters at
strategic points and times during the fall and spring trials of the prototype.
In both cases the waste strength parameter is plotted vs. accumulated fresh
product input to the washers during the trial. All parameters in all trials
tended to follow similar patterns. Data on all trials are contained in
Tables C-16 through C-57, Appendix C.
The six sites in the figures represent sampling locations (refer to
Figures 1 or 12) and differences between adjacent sites in the flow pattern
represent the effect of a certain component of the system on the quality of
the wash water. Values at site 2 minus those at site 1, for instance,
represent the amount of a waste component added to the wash water in washer 1;
similarly for sites 3 and 4 in washer 2. Sites 2 and 6 bracket the effects
of the sump pump and moving-screen belts in sub-system 1; sites 4 and 5 in
sub-system 2. Sites 6 and 1 bracket the effects of the settling tank in the
first sub-system; sites 5 and 3 in the second. Site 1 also represents the
overflow to waste for the entire system.
The most obvious fact from these graphs is the considerable difference
in waste strength in the first sub-system (sites 1, 2, and 6) compared to the
second (sites 3, 4, 5). Approximately 75 percent of the product cleaning
took place in the first washer and settling tank based on analysis of grit
accumulated in the bottom of the tanks and remaining in the water at the end
of each trial. These graphs also indicate the general effectiveness of the
moving-belt screens and settling tanks. The effectiveness of the settling
tanks, as noted earlier, could be improved by lowering the recirculation
rate and eliminating the surges in flow caused by the intermittant operation
of the sump pumps. Also, the settling tanks apparently performed better in
57
-------
250 -
OSITE I
A SITE 2
A SITE 3
• SITE 4
• SITES
O SITE 6
300
200
100
OI23456789
ACCUMULATED PRODUCT INPUT, kgxlC-3
SITE I
SITE 2
SITE 3
SITE 4
SITES
SITE 6
Figure 31: Total suspended solids vs. accumulated product Input
at .-ill six sampling sites. Trial 6, Fall, when processing
col lards with prototype system.
ACCUMULATED PRODUCT INPUT. KG X 10
Figure 32. chemical oxygen demand vs. nrrumuluted product at all six
K.impllng sited. Trial U, Spring, when processing turnip
greens with prototype system.
,200
s
uJ
o
x 100
o
ui
1234!
ACCUMULATED PRODUCT INPUT, KG X IO'
o SITE "2
o SITE 13
• SITE l«
• SITE 15
Figure 33. Total suspended solids v*. accumulated product
at all four sampling Kites, Trial 1, Spring,
apinach processed with conventional washer.
01 23456
ACCUMULATED PRODUCT INPUT, KG X I03
Figure Jti. Chemical oxygen demand vs. accumulated product at
all four sites. Trial 6, Spring, turnip groans
processed with conventional system.
58
-------
the fall trials than the data indicates due to the method of taking samples
at sites 5 and 6 (input ends of the tanks). In the fall trials, water samples
were taken at these points by dipping the container into the surface of the
water. Some rapidly settling solids may have been missed by this technique.
In the spring trials, these samples were taken by holding the container under
the moving-screen belts to catch the samples before the water entered the
tanks.
Ideally, a waste strength parameter in any one of the washers or settling
tanks in a recirculating system should increase by some relationship such as:
x = A(l - e~Bq)
where:
x = concentration of the particular parameter at any time, t.
q = total material processed to time, t.
A and B = constants.
In other words, given constant inputs - i.e., constant rate of material
input of constant "dirtiness" and a constant fresh water input rate — the
waste strength parameter should approach the asymptotic value A with time.
Material input rates, the amount of soil on the vegetables, and even the
range in volatiles produced varied too much to be able to make precise
predictions of concentrations. In general, however, it appears that the
water quality parameters in the prototype system, operated as in these trials,
could be expected to stablize in approximately 5 hours. Leaving the fresh
water on during breaks in processing would, of course, tend to dilute the
waste strength parameters, (note dip in values of TSS for first sub-system
in Figure 31). Very low fresh water input rates as in trial 4 in the spring
(Figure 32) would tend to increase the time until stability is reached.
Accumulations of dissolved organics may also have the effect of lowering the
emission rates of COD and BOD. The decrease in osmotic gradients between
the product and the wash water could reduce the leaching of these materials
as washing proceeds with recirculated water.
Twenty-day BOD values were taken on some of the water samples during the
fall trials. These are tabulated in Tables C-54 through C-57, Appendix C.
Figures 33 and 34 show examples of water quality parameters vs.
accumulated product for the sampling sites on the conventional washers. The
waste strength parameters are again plotted vs. accumulated fresh product in-
put to the washers. Generally the concentrations in the first washer (sites
7 and 8 or 12 and 13) were higher than those in the second (sites 9 and 10
or 14 and 15). This was not always consistent, strongly affected by the
amount of water used in each washer. Waste strength also varied considerably
from beginning to end of the trials and inconsistently with the amount of
product processed. This inconsistency was the result of, not only variations
in incoming product quality, but also in water flow rates.
59
-------
£H
pH was measured and recorded every time water samples were taken.
The range in the fall trials was 6.5 to 7.6 and in the spring trials 7.6
to 8.5. These ranges do not indicate any problems for waste treatment due
to pH.
Color, Turbidity and Conductivity
Readings on color, turbidity and conductivity were taken and recorded
during the fall trials with the prototype. These parameters generally
followed the trends of the other waste strength parameters. These measurer
ments are easy to make and any or all of them may provide simple means for
controlling the operation of recirculating systems in the future. For
example, Figures 35, 36 and 37 present simple regressions for BOD vs.
color for trials 2, 3, and 4 (fall) with collards. Further study on
these general relationships for each commodity appears warranted.
Pesticides
Water samples for pesticide analysis were taken midway and at the end of
each trial of the prototype-at the overflow of washer 1 (site 2) during the
spring trials, tfhese samples were analyzed for Phosdrin, the insecticide
used by Exmore Foods. The results are listed in Table 13 below.
TABLE 13. CONCENTRATION OF PHOSDRIN IN WATER OF FIRST WASHER OF PROTOTYPE
SYSTEM. SPRING TRIALS
Trial
4
5
6
Water /Product
A/kg
2.18
3.99
4.22
Hours of Operation
3
5
4
7
4
Concentration
ppb*
1.45
0.81
0.17
non-detectable
Trace
£0.01)
6.5 non-detectable
Parts per billion.
Samples from Trial 1, the first of the four trials with the prototype
in the spring became overheated in transit, began to decompose anerobically,
and consequently could not be analyzed. Samples from the other trials indicate
that 1) concentrations of pesticide were very low, 2) they tended to decrease
with time rather than build up in the recirculated water and 3) they tended
to decrease with increased water/product ratios.
60
-------
140
130
IZO
110
100
90
SO
7O
60
50
40
30
20
10
Y«0.49X-O.K>
R«0.9I
50 WO 150 20O 250 30O
COLOR IN KLETT UNITS
Figure J5. Five-day biochemical oxygen demand v«. color. Trial 2,
Fall, when processing collards with prototype system.
130
120
110
100
90
80
70
60
50
3O
20
10.
100
90
80
60
S
g 50
g
-J 40
CD 20
10
K) 20 30 4O 50 6O 70 80
COLOR IN KLETT UNITS
Figure 36. Five-day biochemical oxygen demand vs. n.lor. Trl.il 1, tall,
wlicn processing collards with prototype svKten.
10 20 30 40 50 60 70 80 90
COLOR IN KLETT UNITS
Figure 37. Five-day blorhcmlcal nxvgen demand v». color. Trial 4, Fall,
uhen processing collards witli prototype systen.
61
-------
SUMMARY OF WASTE PRODUCTION FROM WASHERS
Waste is carried from the washers by the water in three different
ways: 1) with the effluent during processing, 2) with the water carried out
of the system on the product and 3) with the water dumped from the washers
(and settling tanks for the prototype) at the end of a processing shift.
Tables 14, 15 and 16 show the amounts of each waste parameter measured in
these trials per metric ton of product processed. Concentrations for each
time period multiplied by average water flow during the period were summed
for an entire trial to obtain the total waste leaving in the effluent.
Similarly, the concentrations in the last washer multiplied by average
product flow and amount of water leaving per unit of product were also
summed. The amounts of waste in the washers and settling tanks at the end
of processing was determined by multiplying the final concentrations by the
volume of the tanks in which they were measured.
Tables 15 and 16 (spring trials) show the amounts of waste leaving the
washers by each route. The percentages in each case varied considerably.
The waste leaving the prototype (Table 15) on the product may be slightly
less than estimated here because of the final rinse on the discharge belt of
the second washer. One consistency to be noted for the three trials when
turnip greens were processed (4, 5, 6) is that the percentage of waste
removed by the system overflow increased very rapidly as the amount of fresh
water input increased.
There are at least three things that affect total waste production and
waste stream concentrations. They are: 1) the variety of vegetable being
processed, 2) the amount of water per unit of product processed, and 3)
the condition of the vegetables. Potter (15) shows that collards are high in
nutritive value — 7.5 vs. 4.3 grams of carbohydrates and 40 vs. 26 calories
per 100 grains compared to spinach. This implies that there should be con-
siderable difference in VSS and COD production from different varieties.
On the other hand, Bough (2) found that spinach produced significantly
higher waste loads during washing when compared to collards, turnip greens,
mustard and kale. The results of this study however, do not indicate that
any one of the three products tested could be used as a model for maximum
VSS and COD emission. Emission rates for TSS were consistently higher for
spinach. The combination of savoy leaf surfaces and low growth profile
undoubtedly increases grit accumulation compared to other leafy vegetables.
The amount of water used to wash a unit of product appears to have con-
siderable influence on waste production. VSS and COD mass emission rates
were consistently lower for the prototype than for the conventional system.
There are at least four possible reasons for this: 1) The concentrations
of these waste components in the recirculated water of the prototype were
considerably higher than those in the conventional washer. The more dilute
and larger amounts of water used in the conventional washers may have
affected the surface of the leaves and induced more leaching of organics,
a possibility noted by EPA (7). 2) The water used in the conventional washers
was, for the most part, taken from the product cooling flumes and did contain
some soluble material. 3) A significant amount of these components may have
left the system with the leaf fragments separated out by the moving belt
62
-------
TABLE 14. WASTE LOADS DISCHARGED WITH WATER* FROM PROTOTYPE SYSTEM
DURING FALL TRIALS. 1975
Trial
1
2
4
5
Product Waste Stream/
Product
(A/kg)
Collards 1.97
Collards 0.69
Collards 1.30
Spinach 1.55
Waste Load,
TSS VSS
0.38 0.19
0.27 0.21
0.43 0.24
2.44 0.21
(15.0)**
, kg/metric ton+
COD BOD
0.92 0.25
0.77 0.16
0.88 0.22
0.91 0.17
Sum of wastes carried out overflow of system during trials, plus waste
in water carried out on product, plus waste in water remaining in system
at end of processing.
**
Grit collected from bottom of washers and settling tanks at end of trial.
Amount of grit from this source was negligible in other trials.
Waste loads given in kg/metric ton of fresh (raw) product entering washing
system. See Table 6 for factors to convert these readings to field weight
or packaged weight.
63
-------
TABLE 15. WASTE LOADS DISCHARGED WITH WATER FROM PROTOTYPE SYSTEM DURING
SPRING TRIALS. 1976
Waste stream/ Waste Load, kg/metric ton"*" and
Product Waste* percent of total
Trial Product A/kg
1 Spinach 1.38
4 Turnip 0.16
Greens
5 Turnip 1.35
Greens
6 Turnip 1.68
Greens
Source
A
B
C
TOTAL
D
A
B
C
TOTAL
D
A
B
C
TOTAL
D
A
B
C
TOTAL
D
TSS
0.65(25.4%)
0.97(38.2%)
0.93(36.4%)
2.54(100%)
5.30
0.43(38.1%)
0.06( 4.9%)
0.64(57.0%)
1.13(100%)
2.10
0.19(32.5%)
0.20(34.2%)
0.20(33.3%)
0.59(100%)
1.10
0.06( 6.9%)
0.53(65.6%)
0.22(27.5%)
0.81(100%)
2.90
VSS
0.11(26.6%)
0.15(38.0%)
0.14(35.4%)
0.40(100%)
0.07(44.8%)
0.01( 6.9%)
0.07(48.3%)
0.15(100%)
0.04(38.1%)
0.03(28.6%)
0.04(33.3%)
0.11(100%)
0.01( 8.3%)
0.08(66.7%)
0.03(25.0%)
0.12(100%)
COD
0.37(32.4%)
0.39(34.7%)
0.37(32.9%)
1.13(100%)
0.20(35.2%)
0.02( 3.6%)
0.34(61.2%)
0.56(100%)
0.13(27.7%)
0.15(31.9%)
0.19(40.4%)
0.47(100%)
0;04( 8.2%)
0.32(64.3%)
0.13(27.5%)
0.49(100%)
A - Carried out with water on product.
B - Discharged from settling tank #1.
C - Remaining in water in system at end of processing.
TOTAL - Total of wastes from water in system
D - Total grit collected in the bottom of the washers and settling tanks
at the end of each trial.
+ - Waste loads given in kg/metric ton of fresh (raw) product entering
washing system. See Table 7 for factors to convert these readings
to field weight or packaged weight.
64
-------
TABLE 16. WASTE LOADS DISCHARGED WITH WATER FROM CONVENTIONAL WASHERS DURING
SPRING TRIALS. 1976
Waste stream/ Waste Load, kg/metric ton+ and per-
Product Waste* cent of total
A/kg
Trial
Product
Source TSS
VSS
COD
1 Spinach 17.5
2 Spinach 4.1
3 Spinach 9.8
4 Turnip 17 . 6
Greens
5 Turnip 16.0
Greens
6 Turnip 20.7
Greens
A
B
C
TOTAL
A
B
C
TOTAL
A
B
C
TOTAL
A
B
C
TOTAL
A
B
C
TOTAL
A
B
C
TOTAL
1.74(12.2%)
12.10(85.1%)
0.39( 2.7%)
14.23(100%)
4.30(46.5%)
4.52(48.9%)
0.43( 4.6%)
9.25(100%)
0.81(16.8%)
3.55(74.2%)
0.43( 9.0%)
4.79(100%)
0.24( 9.3%)
2.26(87.4%)
0.09( 3.3%)
2.59(100%)
0.12( 8.1%)
1.24(87.3%)
0.07( 4.6%)
1.43(100%)
0.21(14.6%)
1.14(79.2%)
0.09( 6.2%)
1.44(100%)
0.22(13.0%)
1.37(82.5%)
0.07( 4.5%)
1.66(100%)
0.40(42.0%)
0.51(54.3%)
0.04( 3.7%)
0.95(100%)
0.15(17.2%)
0.62(73.4%)
0.08( 9.5%)
0.85(100%)
0.04(10.3%)
0.29(85.3%)
0.02( 4.4%)
0.35(100%)
0.04(10.4%)
0.29(85.1%)
0.02( 4.5%)
0.35(100%)
0.03(13.0%)
0.19(82.6%)
O.OK 4.4%)
0.23(100%)
0.39( 8.7%)
3.93(87.1%)
0.19( 4.2%)
4.51(100%)
0.65(46.8%)
0.65(46.8%)
0.09( 6.4%)
1.37(100%)
0.56(15.9%)
2.59(74.1%)
0.35(10.0%)
3.50(100%)
0.45(13.5%)
2.74(81.9%)
0.15( 4.6%)
3.34(100%)
0.49(15.1%)
2.57(79.3%)
0.18( 5.6%)
3.24(100%)
0.42(18.9%)
1.62(73.6%)
0.16( 7.5%)
2.20(100%)
A - Carried out with water on product.
B - Discharged from washers 1 and 2
C - Remaining in water in system at end of processing.
TOTAL - Total of wastes from water in system. Grit remaining in bottom of
washers at end of each trial could not be collected without inter-
ferring with plant operations.
+ - Waste loads given in kg/metric ton of fresh (raw) product entering wash-
ing system. See Table 8 for factors to convert these readings to field
weight or packaged weight.
65
-------
screens. 4) Some of these components may have settled out with the finer
soil particles in the prototype settling tanks. Soil samples taken from
these tanks were observed to be high in organic matter.
The amount of grit removed by the water from the conventional system
(TSS minus VSS) appears to be greater than that for the prototype system.
No direct comparisons, however, can be made because the grit collected in the
bottom of the conventional washers could not be measured in these trials.
The prototype showed a consistent advantage over the conventional washers in
reducing the amount of grit on product samples that were hand washed as
described earlier. Assuming that this is a correct representation of the
relative effectiveness of the two systems, then the prototype has the added
advantage of consolidating more of the wastes in its washers and particularly
in the settling tanks for disposal separate from washer effluents.
Emission rates of the various waste components are also strongly
influenced by the conditions of incoming product within a variety. This
condition is influenced by many things including maturity of plants, growing
conditions, whether wilted or turgid, rain or irrigation prior to harvest, etc.
These effects are demonstrated by the data for each variety presented in
Tables 14, 15, 16 and in Table 17 (described below). Even if these data were
"normalized" to constant amounts of water per unit of product processed there
would still be considerable differences in TSS, VSS and COD per kg of product
within each variety.
Table 17 is a summary of average waste stream size and average waste
component concentrations for each of the trials, fall and spring. These data
do not reflect the changing concentration of waste components in the prototype
system with time. They should, however, be useful in planning for design
of a waste treatment system that uses either conventional or recirculating
washers.
Overall average operating conditions for the prototype washing system
during these trials included a product input rate of 1278 kg/hr and fresh
water input of 72 &/min. Under these conditions 2.20 A/kg left the system
with the product, 1.18 left via the waste stream and the waste concentrations
(TSS, VSS, COD) in the various units of the system could be expected to
stabilize in approximately five hours of continuous operation. The number
of trials run were insufficient to indicate whether or not these were
"optimum" conditions. Nevertheless, they appear to be "good" or minimum
conditions for producing suitably clean product. For all product flow
rates a minimum fresh water input of 3.5 A/kg (0.42 gal/lb) is recommended.
ECONOMIC COMPARISONS
The following is an example problem to demonstrate the comparative
economics of owning and operating a two-washer prototype system vs. a two-
washer conventional system of equal output. Many variables will, of course,
affect this type of comparison. The basis here is an assumed "reasonable"
set of operating and economic conditions.
66
-------
TABLE 17. WASTE STREAM CHARACTERISTICS FROM PROTOTYPE AND CONVENTIONAL
SYSTEMS
Average Waste
Stream ±
Trial
1-F*-P+
2-F-P
3-F-P
4-F-P
5-F-P
1-S-P
4-S-P
5-S-P
6-S-P
1-S-C
2-S-C
3-S-C
4-S-C
5-S-C
6-S-C
Product
Collards
Collards
Collards
Collards
Spinach
Spinach
Turnip
Greens
Turnip
Greens
Turnip
Greens
Spinach
Spinach
Spinach
Turnip
Greens
Turnip
Greens
Turnip
Greens
£/min
38.2
20.0
18.9
25.8
25.5
29.5
3.9
24.1
37.7
328.3
107.2
266.3
405.9
396.3
393.0
Avg. Waste Cone.
TSS
80.8
91.7
158.0
121.4
516.8
818.5
414.5
194.4
230.0
1157.1
3618.0
336.7
129.3
70.6
54.9
VSS
30.6
68.6
95.0
65.9
45.6
126.9
48.0
28.0
35.6
103.1
168.9
60.1
17.2
16.4
9.1
.*, mg/Jl
COD
137.6
264.1
346.3
127.7
144.0
331.1
191.7
131.7
139.9
216.2
255.1
228.9
156.2
147.8
78.6
BOD5
46.8
62.6
62.5
57.0
42.3
*
F = fall trial, S =
+
P = pro
to type, C =
spring trial
conventional
= Values are averages for the single waste stream from the prototype
system and are weighted averages for the total waste flow from the
two conventional washers.
67
-------
Annual Fixed Costs
The assumptions here are: 1) The two conventional washers will cost
$12,000 ($6,000 each. Estimate from A. K. Robins Co., Baltimore, Md.) and
the prototype will cost $16,000. The prototype cost is based on an estimate
of 1.33 times that of the conventional washers and assumes that it will be
constructed to include the simplifications cited in the recommendations
(section 3 of this report). 2) Useful life of the equipment is 12 years and
salvage value at the end of usefulness will be 10% of initial cost. Straight
line depreciation will be used. 3) Interest on investment will be equal to
8% of the average value of the equipment over its useful life, per year; and
4) cummulative over ownership costs (taxes, housing, insurance) are equal to
2% of the initial cost per year. The following annual costs are derived
using these assumptions.
Item Washer Line
Conventional Prototype
Depreciation $ 900 $1200
Interest on investment 516 688
Taxes, housing, insurance 240 320
Totals $1656 $2208
Yearly difference: $2208 - 1656 = $552.00
Operating Costs
Operating costs and variables are taken from local information
(Blacksburg, Va.) and conditions comparable to those reported in this study.
Only the waste stream from the washers is considered in computing sewer
charges. Assumptions include:
Item Washer Line
Conventional Prototype
Hours of operation 16 16
Man-hours labor/day 18 20
Labor cost, $/hr 3.5 3.5
Product throughput, Ib/hr 3000 3000
Input water rate, gal/min 100 20
Water for filling system, gal. 1500 3000
Waste stream, gal/min 35 6.5
Waste production., into waste
stream, Ibs/ton of product,
(Assumes that 3/4 of the product
processed is spinach, 1/4 turnip
greens):
TSS 11.53 3.16
VSS 1.49 0.48
COD 5.14 1.33
68
-------
Power to operate washers, kw 3.0
Electricity costs, c/kwh 3.4
Fresh water cost, $/1000 gal 0.5
Sewer charge for water, $/1000 gal 1.0
High strength surcharge rates, £/lb
TSS above 200 mg/Jl 10
COD above 120 mg/Jl 7
Repairs and maintenance,
(0.02% of initial cost/hr.) $/hr. 2.40
14.6
3.4
0.5
1.0
10
7
3.20
Using the above assumptions the following daily operating costs were calcu-
lated:
Item
Electric Power
Water (including two fill-ups/
day)
Sewer Charges
Sewer Surcharges
Repairs
Labor
TOTAL
Washer Line
Conventional Prototype
$ 1.62 $ 7.96
49.50
84.60
14.03
28.40
63.00
$251.15
12.60
12.24
4.38
51.20
70.00
$158.38
Daily difference $251.15-158.38 =$92.77
Under the assumed conditions for this problem, then, the average
annual difference in owning the two types of systems could be recovered in
slightly less than 6 days of operating time.
69
-------
REFERENCES
1. Agricultural Research Service, 1963. "Composition of Foods." Agriculture
Handbook No. 8, USDA, Washington, D. C.
2. Bough, W. A. 1973. "Composition and Waste Load of Unit Effluents from
a Commercial Leafy Greens Canning Operation." Journal of Milk and Food
Technology, Vol. 36, No. 11, pp. 544-553.
3. Buckman, H. 0., and N. C. Brady. 1969. The Nature and Properties of
Soils. 7th Edition, The MacMillan Co., Printed in U.S.A.
4. Carter, L. W. 1970. "A Study of Water Conservation and Reuse at the
Stillwell Canning Co. in Stillwell, Oklahoma." Report prepared for
the Ozarks Regional Commission.
5. Day, P. R. 1965. "Hydrometer Method of Particle-Size Analysis."
Methods of Soil Analysis, Part 1, Amer. Soc. Agron., Madison, Wls.
6. Directory of the Canning, Freezing and Preserving Industries. 1971.
E. E. Judge, and Sons Publishers. Westminister, Maryland.
7. E.P.A. 1976. "Final Effluent Guidelines and Standards for Phase II
of the Canned and Preserved Fruits and Vegetables Processing Industry
Point Source Category." Federal Register. Chapter I. Subchapter N,
Part 407.
8. Frey, B. C. 1973. "Modification of a Leafy Vegetable Immersion Washer."
Unpublished Master's Thesis, Virginia Polytechnic Institute and State
University.
9. Frey, B. C., M. E. Wright, and R. C. Hoehn. 1974. "Modification of a
Leafy Vegetable Immersion Washer." Transactions of the A.S.A.E.,
Vol. 17, No. 6, pp. 1057, 1058, 1059 and 1063, St. Joseph, Mich.
10. Holtan, H. N., N. E. Minshall, and L. L. Harrold. 1962. Field Manual
for Research in Agricultural Hydrology. Agricultural Research Service.
Agricultural Handbook No. 224, Washington, D. C.
11. Lopez, A. 1969. A Complete Course in Canning, 9th Edition. The Canning
Trade, Baltimore, Maryland.
12. Mercer, W. H. 1956. "Canner Foods." Industrial Waste Water Control.
Edited by C. F. Guinham, Academic Press, New York, n. Y., pp. 65-71.
70
-------
13. Metcalf & Eddy, Inc. 1972. Wastewater Engineering, McGraw-Hill Book
Co., New York, N. Y.
14. National Canners Association. 1971. "Liquid Wastes from Canning and
Freezing Fruits and Vegetables." Water Pollution Control Research
Series, 12060 EDK-08/71. U.S. Government Printing Office, Washington,
D. C.
15. Potter, N. N. 1968. "Food and Waste." Food Science, The AVI Publishing
Co., Inc., Westport, Conn., pp. 471-587.
16. Ramseier, R. E. 1942. "The Evaluation of Industrial Wastes in the East
Bay." California Sewage Works Journal, XIV, No. 1, pp. 26-37.
17. Robinson, W. H., Jr., and M. E. Wright. "A Note on Plexiglass H S
Flumes." Water Resources Research. In press.
18. S.C.S. Engineers. 1971. Industrial Waste Study on Canned and Frozen
Vegetables, Interim Report, Contract No. 68-01-0021 for the U.S.
E.P.A., Long Beach, California.
19. Standard Methods for the Examination of Water and Wastewater. 1971.
Edited by AWWA, APHA, and WPCF, 13th Edition.
20. Townsend, C. T., I. I. Somers, F. C. Lamb, and N. A. Olsen. 1956.
A Laboratory Manual for the Canning Industry, 2nd Edition, National
Canners Association Research Laboratories, Washington, D. C.
21. U.S. Department of Agriculture. 1971. Agricultural Statistics.
Government Printing Office, Washington, D. C.
71
-------
APPENDIX A
I
460
440
420
400
380
100'
80
60
40
20
•METER I
-o-METER 2
.METER 3
-METER 4
2 3 4 S 6
OPERXTM6 TIME (HRS.)
Figure Arl: W«t«r flov retee ve. operating tla*, trial 2, Fall, 1975,
whan proceaeing eollarde with prototype waahor.
400
380
_ 360
3 340
80
60
40
20
O—'METER 2
23456
OPERATING THE (HRS.)
Figure A-2i Hater flow retea vi. operating clBa, trial 3, Fall, 1975,
vhan proceeelng collarda with prototype waabar.
1
400
380
3 360
340
80
60
40
20
•*»— METER 3
-O—METER 4
-O—METERS
01234567
TIME (HRS.)
Figure A-3: Water flow rataa va. operating ti*a, trial 4, Fall, 1975,
when procaaalng collarda with prototype ayeteei.
560
540
520
5OO
460
440
420
400
380
80
20
•METER 2
•METER I
012345
OPERATING TIME (HR&)
Figure A-*i Hater flow rataa va. operating tiaa, trial 5, Fall,
1875, whan procaaalng aplnach with prototype ayateai.
72
-------
500
I 400
i
METER 3
METER 4
OPERATING TIME (hrs.)
Figure A-S: Water flow rates v«. operating tine. Trill 4, Spring,
1976. when processing turnip greens with prototype
ayate*.
METER 3
METER 4
HETKR 5
1 2 345 678
OPERATING TIME (hri.)
Fltart A-t: Kntrr flow rates v». r.pcr.itln« tl«!, Trl.il 5, Sprint,
W76, wlicn prooMlng turnip (rtenl with prototyp« v»her.
500
400
300
200
500
400
TOTA1.
UASIIKR 1
• UASIILK 2
1 Z 3 4567 8
OPERATING TIME (hrs.)
Figure A-7i Water flnw rates vs. operating ttae. Trial 6, Spring. 1976,
when processing turnip greens'with prototype washer.
J 4 t 6
OPERATING TIHK (hm.)
Figure A-8t Water overflow reteft fron conventional waahers v*.
operating tine. Trial 2, Spring, 1976, when processing
aplnach on the cast line.
73
-------
-*. WASHER 2
•*-• HASHER I
0 12 34 56 78
OPERATING TIME (hrs.)
Figure A-9: Uacer overflow races from conventlonil washers v«.
operating tine. Trial 3, Spring. 1976, vhen processing
spinach on Che uesc line.
1
s
3456
PPERAT1NG TIME (hrs.)
Ktgtire A-10: U.iter over Dow rates from convent lotm! withers vs.
opcrntlnR tlnu*, Trlnl 4, Spring, 197'«, ulion proccxKl
turnip greens on «*«t line.
500 .
3
-------
2500
2000
1500
1000
500
LUNCH
BREAK
BREAK
3 4
OPERATING TIME (HRS.)
Figure A-13: Product flow rate v«. operating tl««, Trl«l 3, Fall, 1975. whsn proceeding
collards with the prototype systea.
2500
2000
1500
1000
500
BREAK
10000
9000
8000
7000
6000
5000
4000
3000
2000
1000
BREAK
LUNCH
BREAK
01234567
OPERATING TIME (HRS.)
Figure A-14: Accumulated product Input vs. operating tl«et Trial
3, Fall, 1975, uhan processing collards with prototype
•ysteau
3 4
OPERATING TIME, hr
Figure A~15: Product flow race vs. operating tlac, Trial 4, Fall, 197S, when processing collards
wlch the prototype systeau
-------
LUNCH
BREAK
01 2 3 4 5 6 T
OPERATING TIME (HRS.)
Figure A-161 Acciueulated produce Input va. operating elno. Trial «, 10000
Fall, 1975, when proce«nlng collardn with the prototype
HyatCB.
-90OO
5 BOOO
V
2 7000
z
J3 6000
£ 5000
ui 4000
3
u
3000
2000
1000
2500
2000
s
£ 1500
1000
300
BREAK
LUNCH
01234567
OPERATING TIME (MRS.)
Figure A-17i Produce flov rate vs. operating time, Trial 5, Fall, 1975, when ;rcceaaing spinach
with prototype aystee.
LUNCH
M
-
2400
2100
1800
.500
Z
g 1200
t> 900
" 600
300
COSVESTICSAL
(Ease line)
I 23456
OPERATING TIMEIHRS.)
Figure A-18: Accumulated product Input vs. orccnttni; tlnr,
Trial S, Kull, 197S, when pro^OHKlng tiplnarh
with the prototype synti*«.
345
OPERATING TIKE (lira.)
Klgurc A-19: Product flow ratr vs. operating tine. Tria* I, Spring,
1976, vhcn processing spinach.
-------
73(10
6000
4500
3000
1500
34 56
OPERATING TIKE (hrs)
Figure A-20: Accumulated product input vs. operating tiise. Trial i,
Spring, 1976, when processing spinach.
Figure A-21: Product flow rate v». operating tl«c. TrUl 2, Spring
e*st ll"" prol:"*'"« 'P1"*'!" "ith the conventional wuhcra.
12000
g 10500
| 9000
| 7500
8 6000
5 4500
3000
1500
0 12 34 5678
OPERATING TIME (hrs.)
Figure A-22: Accuanilated product Input vs. operating time, Trial 2,
Spring, 1976, when processing spinach with the conventional
washers, east line.
6000
4500
a " 3000
0 1 234 56 7
OPERATING TIME (hrs)
Figure A-24: Accusulated product input vs. uporotlng time. Trial 3,
Spring 1976, when proceaslng spinach with conventional
vaaherst weat Una.
3000
2700
- 2400
=£
* Z100
i
1800
g 1500
900
600
300
0 12345678
OPEKAT1SC TIME (hra)
Figure A-23: Product flow rate vs. operating tlfte, Trial 3, Spring,
1976, when processing spinach with the conventional
washers, west line.
77
-------
2100
1300
1200
900
PROTOTYPE
*— CONVENTIONAL
(Went line)
0 12345(78
OPERATING TIME (hr«.)
Figure A-25: Produce flow rate vs. operating tine. Trial 4, Spring,
1976t when processing curnlp greens.
MHIO
4500
3000
J500
—* CONVFHTIONAl
(Vest line)
34 56
OPKKATING Tint (hr«)
Figure A-26i Accustulated product input v*. operating tlaie, Trial
-------
6 8 10
20 40 60 80 100 200
PARTICLE SIZE IK MICRONS
;oo eoo 800
8 10
20 40 60 80 100
PARTICLE SIZE IM MICRONS
400 (00 800
-J
V£>
figure A-31: Sunuclon percentage, vs. particle size for grit accumulated in prototype systeo.
Trill 1, Spring, 1975 when, processing spinach.
Figure A-32: Summation percentages vs. particle site tor grit accuaulated in the prototype syscci
Trial 4, Spring, 1976, when processing turnip greens.
o HASHER 1
A UASHER 2
c SETTLING TAX* 1
" SETTLING TAKK 2
o UASHER 1
c. WASHER 2
= SETTLINC TANK 1
9 SETTLING TANK 2
20 40 60 80 100
PARTICLE SIZE IN MICIIONS
400 600 800
20 40 60 80 100
PARTICLE S1XK IS MICRONS
400 600 800
Figure A-3J! Suonatlon percentages vs. particle slzo for grit accumulated in the prototype Nvyten,
Trial S, Spring, 1976, when processing turnip green*.
Figure A-34: Summation percent,*^*1 vs. particles for grit accumulated in the prototype system,
Trial 6, Spring, 1°7F>, when processing turnip greens.
-------
APPENDIX B
TABLE 1-1. TOTAL INSECTS OH PRODUCT SAMPLES* OF SPINACH GREENS,
TRIAL 5. PALL. DECEMBER 15. 1975
TABLE B-2. TOTAL INSECTS ON- SAMPLES* OF SPINACH GREENS,
Houre of Operation
Site 1.0 2.0 3.0 4.0 SIt*
1 1 12 7 0 1
31061 3
4 0 10 4 0 4
7
4 10
100 gram aaiplea, replicated valuee.
Houra of Operation
.25
0
0
0
0
0
1
1 2 3
0
0
1
000
000
320
4
2
0
0
1
0
0
567
100
100
100
*100 (in .aaple.
TABLE B-3. TOTAL INSECTS AND PXAGMENT COUNTS ON PRODUCT SAMPLES" OF SPINACH CHEEKS, TUAL 2, SPRING,
MAT 12. 1976
Slte_
Houra of Operation
-LI-
7
9
10
Ina* Fratf In». Fra«. Ine. Pra«. Ina. Frag. In». Fr,
12 2
a 2
11 9
18b 5b
0 1
4 • 5
71° 14° 91 23
50° 5°
8 2
86 27
8° 2°
*100 traB saeplea "Average of Tw> Value* cFrag«encs Inaecta
TABLE B-4. TOTAL INSECTS AND FIAOtENT COUNTS ON PRODUCT SAMPLES* OF SPINACH GREENS. TRIAl 3, SPRING.
Site
12
14
15
.25 1 2 3 3'"
IM. Fran. toa. Frq». Ina. Fran. Ina. Frag. Ina. Frag.
83 55 31 43 30
10 34 54 23 11
20 10 02 10 00
*100 graa aaBplae
80
-------
TABLE I-!. TOTAL IKSECTS AND FRACMEHT COUNTS OB PRODUCT SAMPLES* OF TV1NIF GBEQIS, TRAIL 4, SFUKG, TABLE B-6. TOTAL INSECTS AMD FRAGMENT COBHTS ON PRODUCT SAHPLES* OF TUXNIF CIEEKS, TIIAL 5, StlUG,
JUKE t. 1976 JUHE 10. 1976
00
Hour* of OiMrmtlon
Sit* .25 1 2 3 4 5.5
Ins. Vrat. Iu. Fru. Ins. Frtt. Iu._ PT«K. IBJ, Pr«. Ins. Pru/
4 30 10100001 13
12 70 10 1002011 - -
15 10 21201010 - -
TABLE B-7. TOTAL nStCTS AID FIAOOT COTOTS OR PRODUCT SAMPLES* OF TOEKIP GREEKS. TRIAL 6, SPRING
JUHE 11. 1976
Houn of OMratlon
SIM .25 123 4 5 6
In»? fnaft En. Fr«. ln». Fr«. In.. Fr««. Ini. FcM. Iu«. Fr«a. Iiu. Fra».
1 76 12 51)00 00 4)00
3 40 310120000000
4 00 020110000020
12 20 000000------
Houra of Operation
SLt« .25 1 2345 6.5
Inc. Fru. In** Fr*g. In*. Fr«. In». Fru. In». Fru. In>. Fru. tn«. Fru.
4 00 001110012702
12 20 01022221----
IS 01 00331001----
•lOO gca .«,!..
TABLE B-8. MILLIGRAMS OF GUT PER KILOCRAH OF PRODUCT FOR TRIAL 1, SPRING,
WASHING OF SPINACH GREENS. APRIL 22. 1976
Houra of Op.r>elon
Sit* .25 1.3 2 ) 4 5 7
1 - 4770 - 3510 - 2925 5775
3 - 2475 - 1710 - 1665 2280
4 - 1125 - 930 - 675 645
7 10350 - 9195 - 4740
9 4050 - 4215 - 2355
10 1645 - 2595 - 2085
-------
TABLE B-9. MILLIGRAMS OP GRIT PER KILOGRAM OP PRODUCT FOR TRIAL 2,
TABLE B-10. MILLIGRAMS OF GRIT PER KILOGRAM OF PRODUCT FOR TRIAL 3,
Site
7
9
10
TABLE B-ll.
1
3
4
12
14
15
TABLE B-13.
Site
1
3
4
12
14
15
TABLE B-15.
Site
1
3
arum,. KASH1CIU OF SPINACH GREENS. MAY 12. 1976
Knurs of Operation
.83 3 5 7
6255 22275 4245 3510
1845 6525 1920 1920
1275 3765 1215 1080
MILLIGRAMS OF GRIT PER KILOGRAM OF PRODUCT FOR TRIAL 4,
SPRING. WASHING OF TURNIP GREENS. JUNE 4. 1976
Hours of Operation
25 2 4 5
1080 855 1035 675
675 765 660 480
855 1185 750
615 750 285
375 630 270
MILLIGRAMS OF GRIT PER KILOGRAM OF PRODUCT FOR TRIAL o.
SPRING. WASHING OF TURNIP GREENS. JUNE 11. 1976
Houra of Operation
.25 2 4 6.5
1080 1740 915 840
345 840 660 840
150 330 375 480
1125 750 1935
675 630 945
435 435 S10
TOTAL PLATE COUNT (COLONIES X103 PER GRAM) ON PRODUCT FROM
PROTOTYPE FOR TRIAL 3, FALL, WASHING OF COLLARD GREENS,
NOVEMBER 20. 1975
Houra of Operation
1 3 57
682.0 1227.0 350.0 359.0
509.0 473.0 109.0 682.0
SPRING. WASHING OF SPINACH GREENS. MAY 21. 1976
. _ . . . Houra of OgorutJun
Site .25 j
12 1740 1890
14 1230 1230
15 660 900
TABLE B-12. MILLIGRAMS OF GRIT PER KILOGRAM OF PRODUCT FOR TRIAL 5,
SPRING, WASHING OF TURNIP GREENS. JUNE 10, 1976
Hours of Operation
Site .25 2 4 6.5
1 1005 945 1230 615
3 300 352 1410 451
« 195 300 345 242
12 615 1395 1005
14 300 720 570
15 345 510 465
TABLE 8-14. TOTAL PLATE COUNT" (COLONIES X103 PER GRAM) ON PRODUCT FROM
PROTOTYPE FOR TRIAL 2, FALL, WASHING OF COLLARD GREENS,
NOVEMBER 4. 1975
Houra of Operation
Site 0.25 2 4 6
1 I-7 7-3 3.5 1.7
3 «-l 11.6 54.5 2.7
* 6.0 5.3 11.2 8.3
each value is average of two readings.
TABLE B-16. TOTAL PLATE COUNT (COLOSlEs" X103 PER GRAM) ON PRODUCT FROM
DECEMBER 1. 1975
Houra of Operation
1 160. Ob 140.0 290.0 420.0 150.0 430.0 630.0 180.0
3 180.0 20.0 - 100.0 350.0 10.0 4.0 120.0
4 80.0 520.0 40.0 100.0 650.0 50.0 4.0 70.0
*Gran positive rods.
''single value, other values are averngo of two readings*
'single value, other values are average of two readings.
82
-------
TABLE B-17. TOTAL PLATE COUNT (COLONIES X103 PER CRAM) ON PRODUCT FROM ™LB >-18- TOTAL PLATE COUNT (COLONIES X103 PER CRAM) FOR PRODUCT FROM
PROTOTYPE FOX TRIAL 5, FALL. WASHING OF SPINACH CREENS. PROTOTYPE AND CONVENTIONAL SYSTEMS FOR TRIAL X, SPRING,
Hours of Operation
Site 012)4 st"
1 MO.tf 2090.0C 210. Oc 820. Oc 170. 0* *
3 710.0* 100.0* 230.0C 140.0* 3
4 255.0C 150.0" 1090.0° 110. Oc MO.O* *
7
'Average of nw value.. b Average of three values "Average of four or 10
•ore valuea.
Houra of Operation
.25
164.0
14.7
6.0
155.0
96.4
4
245.0
222.0
52.7
31.6
14.2
7
265.0
182.0
104.0
TABLE B-19. TOTAL PLATE COUMT (COLONIES XIO PER CRAM) FOR, PRODUCT FROM TABLE B-20. TOTAL PLATE COUNT (COLONIES XIO PER CRAH) FOR PRODUCT FROM
CONVENTIONAL SYSTEM FOR TRIAL 2. SPRING, HASHING OF SPINACH CONVENTIONAL SYSTEM FOR TRIAL 3, SPRING, WASHING OF SPINACH
CREENS, HAY 21. 1976
Houra of Operation Houra of operation
Site 2 4
7 35.5 95.5
9 2.1 1.2
10 1.6 0.9
7.5 site .25 2 3
0.6 12 3.3* «.l* 136.0*
0.4 14 0.2* 2.6 1.8"
Z-9 IS 0.3 0.7 2.0°
Average of two values.
TABLE B-21. TOTAL PLATE COONT (COLONIES XIO3 PER CRAM) FOR PRODUCT FROM TABLE B-22. TOTAL PUTE COUNT (COLONIES XIO3 PER CRAM) FOR PRODUCT FROM
CONVENTIONAL SYSTEM FOR TRIAL 4, SPRING, HASHING OF TURNIP PROTOTYPE AND CONVKNT10NAL SYSTEMS FOR TRIAL 5, SPRING,
HASHING OF TURNIP GREENS. JUNE 10. 1976
Site
12
1A
IS
Hours of Operation
.25 1 2 3 5.5 Site
4.1 4.6 5.6 8.7 3.5 1
0.7 4.3 5.6 2.1 2.3 )
157.0 l.l 3.0 2.4 1.5 4
12
IS
.25 1
0.2
0.1
0.1
2.6 3.7
1.2 0.5
2.8 0.5
Houra of Operation.
234
0.4
0.3
0.2
61.0 10.1 2.9
0.8 11.9 3.1
19.5 4.8 3.8
6.5
1.1
0.1
0.2
TABLE B-23. TOTAL PLATE COUNT* (COLONIES XIO PER CRAM) FOR PRODUCT FROM
PROTOTYPE AND CONVENTIONAL SYSTEMS FOR TRIAL 6, SPRING,
HASHING OF TURNIP CREENS. JUNE 11. 1976
Site
1
3
4
12
14
15
.25
2.32
0.7
0.3
20.5
0.4
1.5
Houra of Operation
1234
0.9
0.6
0.4
2.3 15.7 B.I 1.4
9.1 2.8 1.7 1.9
29.3 2.4 6.2 1.6
6
10.7
0.6
0.3
'
*Av«r>ge of two valu*».
83
-------
APPENDIX C
TABLE C-l. TOTAL PLATE COUNT (COLONIES* X 103 PER M1I.LILITER) IN WASH TABLE C-2. TOTAL PLATE COUNT (COLONIES X 103 PER HIU.ILITER) IN WASH
HATER FROM PROTOTYPE FOR TRIAL 1, FALL, WASHING OF COLLARD
WATER FROM PROTOTYPE FOR TRIAL 2. FALL. WASHING OF COLLARD
GREENS. NOVEMBER 4. 1975
Site
1
2
3
4
5
6
1
198.0
0.2
0.2
24.6
13.3
10.2
Hours of Operation
}
10.2
15.0
33.0
6.0
0.5
5.8
4
232.0
>30.0b
3.3
3.9
110.0
710.0
Sice
I
2
3
4
5
6
^Average of
0.2J
15.8C
24.7C
11. Oc
14 .Oc
13.7C
18.0C
tvo values. Average of
J
345.0*
561.0*
2.7b
7.6*
3.2*
1573. Oh
4
1177. 5C
1065.0'
9.2b
3.8*
12.7b
1270. Ob
c
'Average of "ore than one value. blnsufficlent dilutions before plating.
TABLE C-3. TOTAL PLATE COUNT (COLONIES X 103 PER HILLILITER) Ut WASH
•••-—X FROM PROTOTYPE FOR TRIAL 3, FALL, WASHING OF COLLARD
All valuea replicated. Colonies vere 991 gaeillua subtilis.
TABLE C-5. TOTAL PLATE COOT (COLONIES X 10 PER HILLILITER) IN WASH
L WASHING OF SPINAC
TOTAL PLATE COOT C
WATER FROM PROTOTYPE TOR TRIAL 5, FALL, WASHING OF SPINACH
of t«o v,Ou«. "inaufficlent dilution, before plating-
TABLE 0-7. TOTAL PLATE COUNT (COLONIES X 103 PER KILLILITER) IN WASH
u.^. nc rnmENTIONAL SYSTEM FOR TRIAL 3. SPRING, WASHING
'Aver.ga of two valoea.
TABLE C-4. TOTAL PLATE COUNT (COLONIES X 10* PER HILLILITER) IN WASH
WATER FROM PROTOTYPE FOR TRIAL 4, FALL WASHING OF COLLARD
GREENS. DECEMBER 1. 1975
1
4
16.0 100.0
0.1 18.0
7
100.0
32.0
Site
1
2
3
4
5
6
Hours of Operation
137
36.0* 3.9 9.2*
18.0* 7.8* 9.43
2.8*
0.2" 2.9°
0.2 6.6*
16.1 5.3a 14.0
^Average of two values.
TABLE C-6. TOTAL PLATE COUNT (COLON1KS X 10' PER HILLILITER) IN PROTOTYPE
AND CONVENTIONAL WASH WATER, TRIAL 1, SPRING, WASHING OF
1
4
8
10
.25 2 3 4 7
0.3 90.0 25.4
0.2 7.0 5.3
12.0 14.6 700.0
3.6 4.4 1860.0
TABLE C-8. TOTAL PLATE COUNT (COLONIES X 103 PK» MILHLITF.R) IN PROTOTYPE
AND CONVENTIONAL WATER FOR TRIAL 4, SPRING, WASHING OF TURNIP
GREENS. JUNE 4, 1976
Site
1
1
6
13
15
.25
>30.0*
>30.0*
>M.O*
>30.0*
>30.0*
2
-
63.0
85.0
3
200.0
450.0
390.0
_S.J
30.0-
300.0*
600.0
38.0
31.0
•lo.ufdci.nc Dilution.
84
-------
2
4
13
IS
AND CONVENTIONAL WATER
CREEKS. JUNE 10p 1976
.25 3
9,7
2.5
38.0 ZZO.O
44.0 300.0
TOR TRIAL i, SPRING. WASHING OF TUI
Koura of Operation
4 6.5
26.0 24.0
29.0 31.0
1.2 40.0
55.0
39.0
•IIP
1
2
4
13
15
AND COW
CRAS.
1».J
106.0
.
49.0
70.0
ttXTlOHU. WATER FOR TRIAL k, SPUN
JUK. 11. 1916
25.0
SJ.S
7.5
290.0 160.0
130.0 260.0
G. WAUIIK OF TVXVir
23.0
31.0
6.0
.
-
., m,ivn™ mi«T (COLOHIES X 102 PER HILLILITER) IN WASH TABLE C-12. TOTAL COLIFORH COUNT (COLOHES X JO2 PER HILLILITER) IX WASH
TABLE 0-U. TOTAL «£»•«««* jgMW J^ ^^ or ^^ ^"ZESTF £& *"* ?> '*"" """^ "' "^
gp-™- n^M». •>* 101«i WPK». TWVtm.VK ..t i*Zl_ . . . .
. — — —
1
2
6
6
3 «
i.oo »•«"•
0.32 °-<>2
0 »
0.06 S*-0
Houra of Operacioi
Cite 0.25 3
1 0.01 15.60
2 0.31* 11.75s
4 0.11*
6 7.1Jb 39.00
t
6
90.00
..IS*
1.00
S.10*
*Iniu((Ul«nt dilution Ixfote pUtln(.
aAMraga of two value*.
t*f thr^e v.thies.
TABLE C-13. TOTAL GoLlFOKM COUNT (COLONIES X 10* PER HILLlLHCRk 1H UASH
WATER T-ROM PROTOTYPE FOR TRIAL 3, FALL. WASHING OF COLLARD
TA1LE C-14. TOTAL COLIFDM CCVXT (COLONIES X 10* FEI KIU.ILITERJ {« HASH
WATER FROM PROTOTYPE FOR TRIAL 4. FALL. WASHING OF COLUXP
Site I 5 i 1
t 0.10 14.00a 4.00 t
4 - 0.05 Z2.00 3
4
*Averase of three valuaa.
1
336.0*
174 .Ob
1.."
2.0"
114.0
Houra of Operation
J F
J9.0" »1.4b
78.0b 93.6b
2..0b
29.0b
H.O*
7g.0c 127.0b
TABLE C-15. TOTAL COLIFORM COUNT (COLONIES X 102 PER HILLILITER) IN WASH
WATER FROM PROTOTYPE FOR TRIAL 5, TALL, WASHING OF SPINACH
JAver10.00b
2 >10.00b
4
5
6 >10.00b
3 t
>10.00b 3.20
>10.00b Z.40*
>10.00b
14.00
>10.00b 21. M*
COLLARP OKOS. OCTOIER 24. 1975
— Sice 0.25 1
1 10* 7»*
2 36 132
3 »" »a
426
5 «" 1Z«
6 2 64
2
10»*
120
17»
2
10
112
3 »
II )W*
100- 152*
41* 20
11 34
3Ja 40
115* 124
'Average o
f two valuae. Inaufflcianc dllutiona before plating.
'Average of two values.
85
-------
TA"-EC-"-
Site
1
2
3
5
6
Average of
TABLE C-19.
Site
1
2
3
4
5
6
^Average of
TABLE C-21.
1
2
3
4
5
6
7
8
10
*Average of
t r«u^., mummu vr lubLJUUI uiujfcfla, IWVEMBEK «t , iy/3
Houra of Operation
0.25 123456
24* 37 58° 92 92* 119° 220*
32 47* 76 120* 128 155* 180*
20 11 15* 24 29* 22 64
10 16* 18 . 44* 36 SB 72
18* 22 24° 12 44° 60 76
25 43° 76 134° 116 145° 210*
two values.
ANALYTICAL CONCENTRATIONS («g/l) OF TOTAL SUSPENDED SOLIDS,
Hours of Operation
15* 27* 162 69* 184 166 148* 200°
16* 53* 68° 94 230° 184 214° 232*
11 7 15 22 33 47 42 44
12* 10 10 37* 41 56* 46° 36*
67 10 18 31 48 60 48 20
16* 58 63* 96 211* 196 184* 200
two values.
ANALYTICAL CONCENTRATIONS (ng/l) OF TOTAL SUSPENDED SOLIDS,
T«I»I. 1, SPBTNC, WASHING OP SPINACH GREENS, APRIL 22. 1976
025 1 2 3 4 5 6 7
243 637 1057 1213 913 797 750 938
314 1069 1537 1673 533 1130 690 1280
69 98 351 514 475 337 277 318
78 100 371 559 158 369 333 397
347 960 1485 1680 910 1101 827 1104
1321 2188 1252* 199 295
1277 2175 1145° 265 239
449 1219 849* 544 366
445 1156* 680* 471 379
; two values.
TRIAL 3, FALL, WASHING OF COLLARD GREEDS. NOVEMBER 30. 19J5
Site 1357
1 52* 184 240 156
4 26* 44 68 225
6 188
TABLE C-20. ANALYTICAL CONCENTRATIONS* <«g/l) OF TOTAL SUSPENDED SOLIDS,
TRIAL 5, FALL. WASHING OF SPINACH GREENS. DECEMBER 15, 1975
Hours of Operation
Site 01234
1 39 384 720 752 689
2 32 491 802 902 632
3 16 112 199 224 227
4 29 137 217 274 280
5 18 121 207 274 274
6 41 456 766 902 712
Average of two valuea in each case.
TABLE C-22. ANALYTICAL CONCENTRATIONS (•(/() OP TOTAL SUSPENDED SOLIDS,
TRIAL 2. SPRING. WASHING OF SPINACH GREENS. HAY 12, 1976
Hours of Operation
7 2596* 6830* 1354 1047° 465° 953° 950
8 2464* 8306° 1186 996° 455* 1022° 929
9 1760° 3780° 24I4" 1077" 515* 951° 1026
10 1612° 3716° 2195° 1041° 436* 851° 985
°Averago of two values.
TABLE C-23. ANALYTICAL CONCENTRATIONS (mg/f) OF TOTAL SUSPENDED SOLIDS,
TRIAL 3. SPRING, WASHING OF SPINACH GREENS. HAY 21. 1976
Hours of Operation
Site 0.25 123 3.25
12 162 343 469° 523' 446*
13 158 423 304° 490 419*
14 128 356 321 446° 376°
15 121 311 280 449° 416*
Average of two vali»s.
86
-------
TABLE C-24. ANALYTICAL CONCENTRATION (•»/« OF TOTAL SUSPENDED SCUDS,
TABU C-25. ANALYTICAL CONCENTRATIONS («B/D OP TOTAL SUSPENDED SOLIDS,
Houra of Operation
SIC*
1
2
4
5
6
12
13
14
15
TABLE C-26.
0.25 1 2
115 367 399
324* 541 530*
67 144 221
117 228 200
103 212 238
323 513 494*
S3 128 230
62 155 227
53 115 185
54 93 164
ANALYTICAL CONCENTRATIONS (ag/l) OF
TRIAL 6. SPRING. WASHING OF TURNIP
3 4 5.5
440* 588* 578*
476° 608* 635*
212° 275* 326*
258* 2«0* 329*
230* 316* 339*
494* 603" 635*
119 81
166 125
115 79
119 72
TOTAL SUSPENDED SOLIDS,
GREENS, JUNE 11. 1976
Houra of Operation
SUe
1
2
3
4
5'
6
12
13
14
15
'Average of
0.25 1 2 3
138 204 376* 365*
216 290 461° 430*
45 68 129° 134
62 91* 141 142
SO 70 139 137
214* 278 476* 414
41 43 45 79
37 50 52 71
22 32 47 46
23 40 40 59
cvo values.
4 5 6
179 166 162
224 213 232
79 71 89
91 80 92
84 82 98
226 198 220
77
75
59
84
Site 0,25 1 2
1 68 166 153
2 94* 210 200
3 27 75 40
4 27 85 40
5 23 77 90
6 87 152 229
12 31 82 133
13 39 76 148
14 21 52 76
IS 26 40 79
'Average of two values.
TABLE C-27. ANALYTICAL CONCENTRATIONS
TRIAL 1. FALL. WASHING Of
Hours of Operation
3 4 5
410° 274 158
530* 3J3 187
160 149 101
168 110 69
120 ISA 65
47Ia 324 184
94 88
106 89
74 54
59 48
6.5
132
155
68
72
157
154
<«g/») HP VOLATILE SUSPENDED SOU US,
COLLARD GREENS. OCTOBER 24. 2975.
Houra oC Operation
Site 0.25 1
1 X* 17°
2 2 44
3 IS*
422
5 5* 12*
6 28
'Average of two values.
TABLE C-28. ANALYTICAL CONCENTRATIONS
TRIAL 2, FALL, WASHING UF
Hours
Site 0.25 1 2
1 13* 27 41*
2 20 36* 56
3 7 9 15*
4 9 15* 14
5 10° 13 22*
6 14 30* 58
2 3
36* 38
46 50
13* IS*
4 18
18 20*
33' S3'
4
61*
70*
14
34
34
80
<•«/•) OF WILATtLF. SIISPEHIIKD SOLIDS.
COLLARD GREENS. NOVEMBER 4, 1975.
345
68 80' 96*
82* 108 125°
26 25*
29* 32 42
10 39* 46
94* 84 115*
6
155*
145*
40
56
64
135*
Average of two v*luo»
87
-------
TABLE C-29. ANALYTICAL CONCENTRATIONS <•»/!.) OF VOLATILE SUSPENDED SOLIDS, TABLE C-10. ANALYTICAL CONCENTRATIONS (•«/!) OF VOLATILE SUSPENDED SOLIDS,
TRIAL 3. FALL. WASHING OF COLLARD QUEENS. NOVEMBER 20. 1975
Hour* of Operation
Site
1
4
6
•A».
TABLE
1
46'
29'
3
106
44
120
raae of two values.
C-31. ANALYTICAL CONCENTRATIONS
TRIAL 5. FALL. WASHING OF
5 7
120 108
50 150
(•g/t) OF VOLATILE SUSPENDED SOLIDS,
SPINACH GREENS. DECEMBER IS. 1975
Hours of Operation
Site
1
2
3
4
5
6
U
5
4
2'
4
S
«•
1
39
52
13
14
18
45
234
66 61 57
69 66 32
19 17 20
31 25 27
23 29 29
68 54 40
'single values. all others are average of cwo readings.
TABLE C-33. ANALYTICAL CONCENTRATIONS («g/e) OF VOLATILE SUSPENDED SOLIDS,
TRIAL 2. SPRING. HASHING OF SPINACH CREEDS. MAY 12. 1976
Houra of Operation
7
8
9
10
190* 306*
177* 521*
161* 278"
145* 267*
138 156a
160 152"
200° 165*
184* 164*
55 81* 77
65 84* 77
72 90* 85
70 86* 86
iltc
1
2
3
i
5
6
Averaae o
TABLE C-32
0.25
8
6°
I*
2
6'
1
2
17* 124
27* 37°
6 12
10
10 15
20 36*
Houra
3
36*
48
8
33*
21
64
f two value*.
. ANALYTICAL CONCENTRATIONS
-------
TABLE C-36. ANALYTICAL CONCENTRATIONS («g/l> OF VOLATILE SUSPENDED SOLIDS, TABLE C-37. ANALYTICAL CONCENTRATIONS (•(/» OF VOLATILE SUSPENDED SOLIDS,
TRIAL 5. SPRIHC. WASHING Of TURNIP GREENS. JUNE 10, 1976 TRIAL 6. SPRING. WASHING OF TURHIP CTEEHS. JUHE 11. 1976
Site 0
1
2
1
4
:
6
12
1}
14
i:
Average of
TABLE C-38.
Sice
1
2
3
4
5
6
Iliture of Operaclon
.23 1 2 3 * 5 6.5
7 18 17 64" 42 26 22
12° 28 26 77° 45 30 25
4 8 - 40 15 12 12
8 12 38 11 13 13
1 11 11 35 16 12 27
6 14 14 76* 52 28 25
13 20 21 17 16
15 IS 19 19 16
13 25 21 17 16
12 17 20 16 15
cwo value*.
ANALYTICAL CONCENTRATIONS (»g/<> OF CIIKHICAL OXYGEN DEMAND,
TRIAL 1. FALL. WASHING OF COLLARD GREENS, OCTOBER 24. 1975
Houra of Operaclon
0.25 12 34
26 91 136 173 262
33 132 169 207 346
10 33 45 69 103
8 47 57 96 144
8a 45 47 93 132
20 110 132 206 303
Single value* all other* are average of two values.
TABLE C-40. ANALYTICAL CONCENTRATIONS («*/') OF CHEMICAL OXYGEN DEMAND,
TRIAL 3. FALL. HASHING OP COLLAHD CKEENS. NOVEMBER 20, 1975
Sice
1
4
6
a
Average of
TABLE C-42.
Sice
1
2
3
4
5
6
Hours of Operation
1357
149 378 458 400
93 186 194 264
409
ANALYTICAL CONCENTRATIONS" (•«./ 1 ) OF CHEMICAL OXYGEN DEMAND,
TB1AI. S, FALL, HASHING OF SPINACH CREEKS, DECEMBER IS. 1975
Houra of Operaclon
0 12 3 *
25 113 171 200 211
25 115 194 240 246
25 62 79 98 95
16 63 112 129 113
31 59 88 130 115
35 129 181 240 214
Hours of Operation _.
1 25 49 56° 46'' 28 23 22
2 39 73 102'1 55° 29 25 28
3 9 17 23'1 22 15 10 14
4 14 21° 23 24 15 14 15
5 11 17 22 23 IS 16 15
6 41* 85 64* 54° 26 25 25
12 8 7 9 10 9
13 8 8 8 11 10
14 10 11 9 11 9
15 9 10 8 8 11
"Average of two values.
TABLE C-39. ANALYTICAL CONCENTRATIONS <«g/t) OF CHEHICAL OXYGEN DEMAND,
TRIAL 2. FALL. HASHING OF COLLARD GREENS. NOVEMBER 4. 1975
Hours of Operation
SICe 0.25 123456
1 45 92 184 304 355 430 439
2 4» 129 214 349 453 492 428
3 25 35 59 98 131 172 103
4 33 41 76 123 123 209 219
5 31 41 75 117 148 201 232
6 47 102 192 355 394 414 557
Average of two values in cncli case.
TAB1.E C-41. ANALYTICAL CONCENTRATIONS (nig/O OF CHEMICAL OXYGEN DEMAND,
TRIAL 4, FALL. HASHING OF COLLARD GREENS. DECEMBER 1. 1975
Hours of Operacion
1 12 24 86 96 204 244 275 292 321
2 26 140 158 234 290 315 359 367
3 16 20 36 192 67 81 87 106 112
4 28 58 60 89 108 110 131 230
5 2 24 40 66 91 87 117 127 127
6 29 42 238 220 2f>5 290 330 341
"Average of cwo valuea In each cane.
"Average of two values in ««ch cai*.
89
-------
TABLE C-43. ANALYTICALCONCENTRATIONS CM,.) OF CHEMICAL OXYGEN DEMAND. TABLE c.«. ANALYTICAL CONCENTRATIONS („/£) OF CHEMICAL OXYGEN DEMAND.
Hour* of Operation
sic.
l
2
3
4
5
7
8
9
10
*Av«r»M of
TABLE C-46.
Slt«
1
2
3
4
5
6
12
13
1*
IS
*Av*raie of
0.25 1
2
3 4
5 6 7
82* 161' 313 433* 463* 431* 388* 378*
89* 194* 354 579 765* 538* 509* 464
47* 106 76 173 220 170* 132 136
63 124 136* 203* 347* 185* 164* 168*
109 72* 105 183 229 181 145 152
274 506 271* 49 164*
261* 514 204 53 168*
117 247 97* 69 155
113* 27'* Oft •** I1**
Site
Hour* oC Operation
7 280* 557° 154° 279* 113a 221° 210°
8 222 724 139 249 87 208 202
9 277* 386* 267* 247* 128 192* 205*
10 236 385 301 274* 111 149 190*
*Av*rag« of tvo value*.
TABLE C-4S. ANALYTICAL CONCENTRATIONS (»g/() OF CHEMICAL OXYGEN DEMAND,
TRIAL 3. SPRING. HASHING OF SPINACH GREENS. MAY 21. 1976
ANALYTICAL CONCENTRATIONS («s/l> OF
TRIAL 4. SPRING. HASHING OF TURNIP
0.25
60
77*
43
45*
47
87*
60*
63
67*
61
two v«lu«».
x
141
152*
90
82*
94
141*
141
145
192
196
2
163
200*
102
108*
114
190*
185
188
263*
259
CHEMICAL OXICEN DEMAND,
GREENS. JUNE 4. 1976
3
243
269*
157
151*
157
256*
167
161
233*
237
4 5.3
255 288
282* 316*
149 204
159* 205*
157 208
283* 306*
137
125
190*
190
Site
12
13
14
15
AAverafte of
TABLE C-47
Sit*
1
2
3
4
5
6
12
13
14
15
.25
80*
85
82*
71
1
237
236
219
222
«
Houri of
2
252*
190*
213°
186*
Operation
3 3.25
309* 304*
321 349
308* 375°
301 329
two value*.
. ANALY
TRIAL
0.25
32
28
19*
16
16*
44
65*
49*
40*
41
TICAL CONC1
5 SPRING
78*
102
43*
18
38*
98
88*
69
133*
92
ENTRAT:
[ON (M/t)
OF CHEMICAL OXYGEN DEMAND,
. WASHING or TURNIP GREENS. JUNE 10. 1976
155*
157
106*
63
60*
149
199*
200
295*
235
llouri of
216*
220
122*
118
119*
243
193*
184
225*
222
Operation
151* 135 135*
178 149 153*
72* 61* 67
82 67 63
87* 55* 159*
169 145 141
186*
180*
228*
204
*Avirag* of two valuu.
90
-------
TABU C-4H. ANALYTICAL CONCENTRATION <>g/t) or CHEMICAL OXYGEN DEMAND,
TABU C-49. ANALYTICAL CONCENTRATIONS (.g/r) OF FIVE-DAY BIOCHEMICAL
.
Site
1
2
3
4
5
6
12
14
15
°Average ol
TABLE C-50.
Site 0
1
3
4
5
6
*Avera,ge of
TABLE c-53.
Site
1
2
3
4
5
6
. .- '. '~i..-.;: ..~.\..r.:::~.' -.::~~r^::~.~.~~~
. Hours ul ojiftatlon
92* 185* 194° 178* 114* 104* 112*
144 240 228* 200 138* 120 124
38* 68* 92 98* 61° 56° 65°
44 80 104 96* 66 57* 70
47* 79* 122* 90* 66* 55* 72*
136 240 228 200 124 116* 130*
36* 64* 93* 89° 115*
35° 86* 133* 112* 157*
32 76 92° 52 166°
I two values.
ANALYTICAL CONCENTRATIONS (ng/1) OF FIVE-DAY BIOCHEMICAL
NOVEMBER 4, 1975
Hours of Operstlon
i.25 12 3456
8 24* 44 62 95* 108 97
4 8 11* 24° 28 35 20*
3 5 12 25 18 35 59
4 6 8 21° 27° 47 46
10 21 50 87 114 98 91*
two values, other values are average of three.
ANALYTICAL CONCENTRATIONS (»g/l> OF FIVE-DAY BIOCHEMICAL
DECEMBER 15. 1975
Hours of Operation
1234
30b 49b 44b 46b
30b 52C 57b 39°
,* 24b 25C 22*
8* 21" 3ie 22*
21b 47b 51C 32C
SiC* 1 3 1 /
1 28 31 47b 81a
2 » 34b 5«b 89
3 ' 8 16b 2ob
4 10 15* 23* 36*
5 '8b 10 23b 36*
6 30* 17 57° 75°
a b
OXYGEN DEMAND, TRIAL 3, PALL, WASHING OF COLLARD (.KEENS,
NOVEMBER 20. 1975
Hours of Operation
Site 1 5 7
1 48a 77a
4 22 38 41
Average of two values.
TABLE c-52. ANALYTICAL CONCENTRATIONS (ng/i) OF FIVE-DAY BIOCHEMICAL
OXYGEN DEMAND, TRIAL 4, FALL, WASHING OF COLLARD GREENS,
DECEMBER 1, 1975
Hours uf Operation
Site 0.25 12345 67
1 3 27b 50b 60b 77* 75a 88° 76b
2 4* 31C 52b 71C 87b 9lc U6C 97b
3 5 12* 14* 19a 20° 22° 30° 35°
4 2 14° 15a 22° 19" 30a 36° 39°
5 6a 9° 20° 27 26 39 42 38*
6 8° 29b 53C 72b 77° 99C 101C 24
"Average of two values. bAverage of three velues. cAversge of four values.
'Average of two values. "Average of throe valuea. cAverage of four values.
91
-------
TABLE C-S4. ANALYTICAL COWENTIIATIOHS (•»/!> OF 20-DAY BIOCHEMICAL OXYGEN
DEMAND, TMAL 1, FALL, nASHDK OF COUJUU) GUENS, OCTOBER 24,
197$
Site
1
2
3
4
5
6
Hours of Operation
4.0
99
105
10
21
3t
93
TABU C-55. ANALYTICAL OMCtKTIATIOKS (it/I) OF 20-DAY BIOCHEMICAL OXYGEN
DEMAND, TUAL 2, FALL, WASHING OF COLLAltD CHEEKS, NOVEHBEt 4.
1975
Hours of Operation
247a
TABU C-54. AMLYTICAL COMDRMTiaHS (•*/!) OF 20-DAY BIOCHEMICAL OXYGEN
DEMAND TUAL 4. FALL. HASUDIC OF COLLA«D CUKHIS. DECEMBE» 1. 1975
Sit*
1
2
3
4
S
6
Hoars of Operation
0.25 J.O
35b
12 SOC
IS 10*
13"
6"
0 37b
*Av*M|* of tvo values. Avsraga ot three values. ''Average of four valuea.
TABLE C-57. AHALYTICAL COMCSmATIONS (s«/t) OF 20-DAY BtOOIEHICAL OXYGEK
n. T»IAL 5. FALL. HASH1MC OF SFlHAqi CKKOIS. DEC.. 15. 1975
Hours of Operation
SIM
1
2
3
e
5
•
0
la"
U*
12*
«•
17*
13'
1 2
3Se «lk
37" S0e
6* Ji*
13* 30b
12* 27*
3lb tle
3
«9b
*4b
30C
43e
«3e
"C
4
3»c
23b
21*
34*
34*
62e
'Average of tvo values. ""Average of three values. 'Average of four values.
92
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-77-135
3. RECIPIENT'S ACCESSION>NO.
4. TITLE AND SUBTITLE
Minimization of Water Use in Leafy Vegetable Washers
5. REPORT DATE
July 1977 issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Malcolm E. Wright
Robert C. Hoehn (Civil Engr. Dept. - VPI & SU)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Agricultural Engineering Department
Virginia Polytechnic Institute & State University
Blacksburg, VA 24061
10. PROGRAM ELEMENT NO.
1BB610
11. CONTRACT/GRANT NO.
S-802958
12. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Research Lab - Gin., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final 5/1/74 - 1/31/77
14. SPONSORING AGENCY CODE
EPA/600/12
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This project was undertaken to construct and test an improved leafy greens washing
system employing water recirculation, to characterize the quality of the wash water
and waste stream and to make comparisons to conventional washers. The prototype
system produced a cleaner product while reducing water requirements and consolidating
waste loads. The prototype system consisted of two drum immersion washers in series,
each with associated moving belt screens, settling tanks and water recirculation
systems. Construction was similar to conventional washers but with modifications to
improve removal of floating trash and increase hydraulic agitation of product. The
prototype was tested in a commercial processing plant during the fall and spring
harvesting seasons, 1975-76. Sixty-seven metric tons of collards, spinach, and
turnip greens were processed through the prototype in 52 hours of actual operating
time. Conventional washers were monitored for 27 hours (38 tons) for comparison.
Insect and bacteria counts, COD, TSS, VSS, arid several other water and product
parameters were measured at predetermined times and locations. Data were obtained
to predict expected waste loads from the products processed. Wastewater discharge
from the prototype was approximately 1/12 that of the conventional washers.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
Food Processing, Circulation, Canneries.
Freezers, Water Quality
Leafy-vegetable process-
ing, Process modification
Washing systems, Water
reuse
13/B
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
105
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
93
•ft II. S. GOVERNMENT PRINTING OFFICE 1977-757-056/61,99 Region No. 5-11
------- |