EPA-600/2-77-118
June 1977
Environmental Protection Technology Series
                  MEMBRANE  PROCESSING  OF COTTAGE
                                                CHEESE  WHEY
                                     Industrial Environmental Research Laboratory
                                           Office of Research and Development
                                          ULS, 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-H8
                                           June 1977
                MEMBRANE PROCESSING OP

                  COTTAGE CHEESE WHEY
                          By

                    Robert R. Zall
              Department of Pood Science
New York State College of Agriculture and Life Sciences
                  Cornell University
                Ithaca, New York  14853
                Project No. 12060 DXP
                   Project Officers

                    Max W. Cochrane
         Industrial Pollution Control Division
Industrial Environmental Research Laboratory-Cincinnati
               Corvallis, Oregon  97330

                     Larry Dempsey
         Industrial Pollution Control Division
     Industrial Environmental Research Laboratory
                Cincinnati, Ohio  45268
     INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
          OFFICE OP RESEARCH AND DEVELOPMENT
         U.S. ENVIRONMENTAL PROTECTION AGENCY
                CINCINNATI, OHIO  45268

-------
                         DISCLAIMER
This report has been reviewed by the Industrial
Environmental Research Laboratory,  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.

-------
                          FOREWORD
     When energy and material resources are extracted,
processed, converted, and used, the related pollutional
impacts on our environment and even on our health often
require that new and increasingly more efficient pollution
control methods be used.  The Industrial Environmental
Research Laboratory—Cincinnati (IERL-CI) assists in
developing and demonstrating new and improved methodologies
that will meet these needs both efficiently and
economically.

     ''Membrane Processing of Cottage Cheese Whey" was a part
of the Industrial Pollution Control Division's program to
develop and demonstrate new technology for the treatment of
industrial wastes.  A full-scale whey processing plant using
membranes was  constructed to process 300,000 pounds per day
of cottage cheese whey.  The two-step system uses
ultrafiltration (UF) and reverse osmosis (RO) according to a
design previously demonstrated in the Phase I portion of
this project.   Though the operation of the plant is both
easy and straight-forward, the system is industrially
difficult to clean and sterilize with the specific enzyme
cleaners and sanitizers used in this project.

     For further information, contact the Food and Wood
Products Branch of the Industrial Environmental Research
Laboratory — Cincinnati .
                      David G. Stephan
                          Director
        Industrial Environmental Research Laboratory
                         Cincinnati
                             1X1

-------
                           ABSTRACT
 A full-scale  whey  processing  plant using membranes was
 constructed  to  process  300,000  pounds per day of cottage
 cheese  whey.  The  two-step  system uses ultrafiltration  (UF)
 and  reverse  osmosis  (RO)  according to a design previously
 demonstrated  in the  Phase I portion  of this project and
 reported  in  Water  Pollution Control  Series  12060 DXF 07/71.

 Proteins  and  lactose can  be separated from  whey with
 semipermeable cellulose acetate membranes as molecular
 sieves  and a  driving force  supplied  by centrifugal or
 positive  displacement  pumps.

 Biochemical  oxygen demand (BOD) of cottage  cheese whey  can
 be reduced from about  40,000  rag/1 to approximately 2,000
 rag/1  in liquids discharged  from the  system.  The RO permeate
 water can be  used  to clean  the  whey  plant,  providing
 biological contaminants are controlled in the UF-RO plant by
 improved  RO hardware and  chemical sanitizers.

 Though  the operation of the plant is both easy and
 straight-forward,  the  system  is industrially difficult  to
 clean and sterilize  with  the  specific enzyme cleaners and
 sanitizers used in this project.

 Preliminary economic data for 5 months of operations suggest
 that the overall operating  costs will approximate the Phase
 I projects of $220,000  per  year, with some  wide differences
 in cost categories such as  cleaning  expenses.  Subsequent
 data with cost  figures  of a year's operations in a 1973.74
 period showed that cost excesses over original estimates
 were cleaners at $26,000  per  year rather than $3,200 and
membrane replacement at $78,600 rather than $39,000.   Future
 annual reports  will  indicate  better  the real impact of
operating costs on whey processing when the system achieves
prolonged stability.
This report was submitted in fulfillment of Grant number
12060 DXF by Crowley Foods, Inc. under the sponsorship of
the U.S  Environmental Protection Agency.  The report covers
a period from June 21, 1972 to December 1974,  and work was
complete as of April 10, 1975-
                             IV

-------
                          CONTENTS
                                                    Page
Foreword                                             iii

Abstract                                              iv

List of Figures                                       vi

List of Tables                                        ix

Acknowledgments                                        x

Sections

I     Introduction                                     1

II    Conclusions                                      5

III   Recommendations                                  6

IV    Phase I:  Pilot Plant                            7
V     Phase II:  300,000-Pound-Per-Day Whey Plant     13

VI    Plant Design                                    2M

VII   Plant Operation                                 30

VIII  Process Economics                               58

IX    Membrane Replacement of UF Tubes                64

X     Pertinent Publications                          67

XI    Glossary                                        68

XII   Appendices                                      71

-------
                           FIGURES


Number                                               Pag.il

   1    Osmosis and Reverse Osmosis                      8

   2    Mechanism of Membrane Separation                10

   3    Membrane Process for Whey Treatment-Flow
       Schematic                                      11

   4    Whey Plant Simplified Flow Plan                 14

   5    Overall View of 300,000 Pound-Per-Day UF-RO
       Whey Processing Plant During Installation.
       Crowley Foods,  Inc., LaFargeville,  New York.
       April 1972.                                    16

   6    Views of Ultrafiltration Section  and Complete
       Control Panel for 300,000 Pound-Per-Day UF-RO
       Whey Processing Plant.  Crowley  Foods, Inc.,
       LaFargeville, New York.  April 1972.           17

   7    End  View of Sanitary Whey Ultrafiltration
       Cabinet and Feed Circulation Pump.   Crowley
       Foods,  Inc., Largeville, New York.
       April 1972.                                    19

   8    View of Reverse Osmosis Section  of  300,000
       Pound-Per-Day UF-RO Whey Processing Plant.
       Crowley Foods,  Inc., LaFargeville,  New York.
       April 1972.                                    21

 ,  9    View of Reverse Osmosis Pump and  Interstage      \
       Tank of 300,000 Pound-Per-Day UF-RO Whey
       Processing Plant.  Crowley Foods, Inc.,
       LaFargeville, New York.  April 1972.            22

 10    Simplified Flow Diagram of Ultrafiltration
       System                                          26

 11    Simplified Flow Diagram of Reverse  Osmosis
       System                                          28

 12    Front View of Cultured  Products Plant  (Whey
       Building  Located  at  Lower  Extreme Left).
       Crowley  Foods,  Inc.,  LaFargeville,  New York.   31

                           vi

-------
lumber                                                Page

  13   Rear View of Cultured Products Plant  (Whey
       Building Located at Lower Right).  Crowley
       Foods, Inc., LaFargeville, New York.           32

  14   Ultrafiltration-Reverse Osmosis Building.
       Crowley Foods, Inc., LaFargeville, New York.   33

  15   Typical pH Values of Whey Supplied to the
       UF-RO Plant:  High, Average, and Low  Values    34

  16   Typical Acidity of Whey Supplies to the UF-RO
       Plant:  High, Average, and Low Values          35

  17   Typical Total Solids Concentration in Whey
       Supplied to the UF-RO Plant:  High, Average,
       and Low Values                                 36

  18   Ultrafiltration—Permeate Flux During 1972:
       High, Average, and Low Values for Month
       Periods                                        43

  19   Ultrafiltration—Permeate Flux During 1973:
       High, Average, and Low Values for Month
       Periods                                        44

  20   View of Whey Feed Silos and Interstage Tanks
       with Portion of Two Reverse Osmosis Cabinets
       and RO Booster Pump Showing.  Crowley Foods,
       Inc., LaFargeville, New York. April 1972.      46

  21   Reverse Osmosis—Permeate Flux During 1972:
       High, Average, and Low Values for Month
       Periods                                        51

  22   Reverse Osmosis—Permeate Flux During 1973:
       High, Average, and Low Values for Month
       Periods           '            -                 52

  23   RO Permeate Quality During 1972 as Measured by
       6005:  High, Average, and Low Values  for
       Ten-Day Periods                                53

  24   RO Permeate Quality During 1973 as Measured by
       Total Solids Where \% Approximates 7000 MG/L
       6005:  High Average, and Low Values for
       Ten-Day Periods                                54
                           VI1

-------
Number

  25   Characteristics of a Single Re-Membraned RO
       Cabinet During Fall 1974:   High,  Average,  and
       Low Values for Ten-Day Periods                 57
                        Vlll

-------
                           TABLES
Number                                               Page

   1   Approximate Gross Average Composition of Whey
       Products                                        2

   2   Selected Microbiological Data of Products
       Processed in the Ultrafiltration Section       39

   3   Typical Microbiological Data of Products
       Processed in the Reverse Osmosis Section       48

   4   Selected Data to Illustrate Bacteriological
       Results in RO Concentrate and Permeate         55

   5   Dollar Costs for Setting Up a 300,000 Ib/day
       Whey Plant                                     58

   6   Five Months Projected and Actual Operating
       Costs for 300,000 Ib/day Whey Plant            59

   7   Whey Plant Cleaning Supplies Usage             60
       *
   8   Operating Costs during May 1973-April 1974     62

   9   Product Flow Summary of UF-RO Operation,
       May 1973 to April 1974                         63

  10   Cleaning Supplies for the UF-RO Plant,
       May 1973 to April 1974                         63
                           IX

-------
                       ACKNOWLEDGMENTS


The support of the project by the U.S. Environmental
Protection Agency and the help of H.G. Keeler, and Max W.
Cochrane and Larry Dempsey (the Federal Project Officers),
R. J. Burm, and Allyn Richardson is acknowledged with
sincere thanks .

Dr. David J. Goldstein (Abcor) was in large part responsible
for the design and construction of the plant.  He labored
long and hard in carrying out his assignment.

Dr. William Eykamp, J. Tom Selldorf, and Tom Ammerlaan
(Abcor) were very helpful and solved many pressing problems.

James Bender, Research Director for Crowley Foods, Inc.,
deserves special thanks for his devoted labors to the
project.  Mr. Patrick Crowley was in charge of plant
operations since its pilot plant stage.'  Mr. J. Elmer
Crowley, more than anyone, has carried the load and has kept
the project moving forward.  Much support to the project was
given by Mr. William Burtis, Crowley's project finance
officer, and Mr. Charles Carlson, their onsite division
manager .

Dr. Robert R. Zall, Cornell University, served as project
director and prepared this report with the aid of data
extracted from the Phase I document.

-------
                          SECTION I

                        INTRODUCTION
This document report Phase II of the two-part acid whey
treatment demonstration grant to abate pollution from cheese
making.  The offer and acceptance of the Federal grant  for
research and development began in September  1969.  Phase I,
covering a sizable pilot-plant operation lasting 13 months,
was described in the Water Control Research  Series 12060 DXF
07/71 entitled ''Membrane Processing of Cottage Cheese Whey
for Pollution Abatement."

Much of the introduction that appeared in the first document
is repeated in this section because the scope and purpose of
the project remains the same.

Phase II was dependent upon the success of Phase I work and
mutual agreement between the grantee and EPA (then the
FWPCA) to proceed to build the 300,000-pound-per-day whey
treatment plant at LaFargeville. New York.

The pollution problems that existed before 1969 have not
diminished but in fact are even more critical nationwide.
Because whey production is a function of cheese making, more
whey is now being generated because of demands of increased
cheese production .

The status of the pollution problem resulting from the
disposal of acid cheese whey is such that new methods must
be used to deal with it effectively.  This section of the
report presents a summary of the size and nature of the
waste problem and provides a general description of the
membrane, separation process under demonstration..

-------
BACKGROUND AND NEED FOR  PROJECT
The  manufacture of cheese  from either whole  or  skim milk
produces,  in addition  to  the cheese itself,  a
greenish-yellow fluid  known as whey.  Whole  milk is used to
produce  natural and  processed cheese such  as Cheddar, and
the  resulting fluid  byproduct is called  sweet  whey, with a
pH in  the  range of 5  to  7.   Skim milk is  the starting
material for cottage  cheese and gives a  fluid  byproduct
called acid whey, with a  pH in range of  4  to 5-  The lower
pH is  a  result of the  acid developed during  or employed for
coagulation .

Each pound of cheese  produced results in  5 to  10 pounds of
raw  fluid  whey.  The  high  organic content  of whey leads to a
severe disposal problem.   Over 10% of the  nutrients from
skim milk  show up in  acid  whey, including  protein and
lactose.  These materials,  if properly recovered, could
provide  useful products.   Table 1 shows  typical compositions
of whey  and dried whey solids.
    Table 1.   APPROXIMATE GROSS AVERAGE COMPOSITION OF WHEY PRODUCTS
                      IN PERCENTAGES BY WEIGHT
Product
     Nitrogenous
Water     matter
  Fat
Lactose
Acid
Ash
Raw cheese
  whey

Condensed
  whey

Dried whey
93.914   0.7-0.9


50-60   7-0-8.0

 2-6   12.0--14.0
0.05-0.6   4.5-5.0 0.2-0.6   0.5-0.6


 1.0-2.0   28     1.0-3.0   5.0-6.0

 0.3-5.0  65.0-70.0 2.0-8.0   8.0-12.0
         Source:  Proceedings of Whey Utilization Conference,
              June 2-3, 1970.  University of Maryland.

-------
£c.oductign_gf _Cottage_Cheese_Whey.

Some statistical information on the production of cottage
cheese (acid) whey is summarized here to provide a basis for
identifying the needs for acid whey treatment.  In
categorizing industries in the United States, cottage cheese
is considered to be part of fluid milk rather than cheese
production.  Thus, statistics on acid whey are somewhat more
difficult to obtain than those for cheddar and other sweet
wheys.  However, the general trends in the fluid milk
industry are the sameas those for the natural cheese
industry; i.e., the industry is growing, but the number of
plants is decreasing.  These trends will probably still
continue toward exceptionally large regional plants with
distribution over wide areas (4,6).

By 1972, overall production of cottage cheese curd and
creamed cottage cheese had risen to an adjusted value of 1.4
billion pounds.  The acid whey output accompanying this
value was set at 7 billion pounds, based on 5 pounds whey
per pound of cottage cheese.  A present output of at least 7
billion pounds of acid whey per year is considered to be a
conservative value, based on the current rate of growth of
cottage cheese production and on the fact that statistics
for other related acid cheese wheys (bakers, farmers, cream)
are not included.
The_Cgttage_Cheese_Whey__Disp_gsal_Prgblem

The organic nutrients of whey, which go unused, place a
costly burden on sewage systems and waterways.  The BOD of
whey ranges from 32,000 to 60,000 ppm.  About 90% of this
BOD is due to the lactose.  Specific BOD values for cottage
cheese wheys are between 30,000 and 45,000 rag/1, depending
primarily on the specific cheese making process used.

Every 1,000 gallons per day of raw whey discharged into a
s.e_wage treatment. £lant can impose a load equal to that from
1,800 people.  This is partially passed into streams in most
cases because BOD removal is not complete.  Every 1,000
gallons of raw whey discharged into a stream requires for
its oxidation the dissolved oxygen in over 4.5 million
gallons of unpolluted water.

-------
Recent statistics give BOD's of about 0.2 pounds per pound
of cottage cheese curd.   Combining this with the production
statistics given above,  a total BOD removal of over 280
million pounds is required for complete waste treatment.
Considering that at the  very best only half of the cottage
cheese whey produced is  currently put to good purpose and
that curd wash water can contain 3% solids, the disposal
problem is severe.

-------
                         SECTION II

                         CONCLUSIONS
Membrane processing of cottage cheese whey can be used on an
industrial scale to capture proteins and lactose fractions.
The BOD of the raw whey can be reduced by this method as
much as 95£ from about 40,000 rag/1 to approximately 2,000
rag/1 .

Ultrafiltration and reverse osmosis membrane plants are
physically easy to operate.

Whey must be produced, handled, and treated much like
pasteurized dairy products with methods that avoid microbial
contamination.  No pretreatment is necessary.

RO and UF membrane technology need improvement in both
operability and durability.

Membranes fail faster when overly exposed to or stored in
sanitizing solutions.  Chlorinated sanitizers appear more
harsh to membranes than quaternary ammonium compounds.

The life expectancy of UF membranes in Phase II appears to
be about half the time that was projected in Phase I of this
project.  Membranes are durable for approximately 1 year
rather than 2 years.

The life expectancy of RO membranes is not clear at this
time.  Data still suggest they can last 2 years, as was
projected from the Phase I pilot study.

-------
                         SE.CTION III

                       RECOMMENDATIONS


Crowley Foods, Inc. recommends that this process be further
improved for industry use as follows:

1.   Future installations should have single-task plumping
    with small hydraulic loadings to simplify both running
    and clean-up functions.

2.   Reverse osmosis hardware should be further simplified,
    and the ultrafiltration system should possess more
    strength and durability than experienced in the
    Phase II plant .

3.   Cleaning costs are believed to be too expensive.
    Additional research is needed to develop both methods
    and improved substances for cleaning and sanitizing
    membranes .

4.   More emphasis should be placed on water treatment
    equipment for plants using membrane technology
    because not all water supplies are compatible with
    membrane systems.

-------
                         SECTION IV

                    PHASE I:  PILOT PLANT
A two-step membrane separation process was developed for the
treatment of cottage cheese whey.  Although the
demonstration has been for acid whey treatment, the
technology is applicable to sweet wheys.  In this process,
whey is simultaneously fractionated and concentrated to give
protein and lactose byproducts.  The final effluent has a
BOD approximating 2000 mg/1, mostly lactic acid, and can be
reused within the cottage cheese plant.  One application,
for example, is in curd washing, another to serve as wash
water.  If the effluent is discharged, a final treatment for
residual BOD removal may be required, depending on local and
state regulations.

The two-step whey treatment process is based on the
application of ultrafiltration (UF) and reverse osmosis
(RO).  Reverse osmosis has been extensively studied during
the past ten years under funding from the Office of Saline
Water; the bulk of this work focusing on desalination of
brackish and sea waters.  Additional publicity has been
given to its application to whey concentration by the United
States Department of Agriculture.  Ultrafiltration is a
variation of this membrane separation technique.

Figure 1 shows the basic concept involved.  A semipermeable
membrane separates water and a solution.  In the absence of
a hydrostatic pressure differential, water will permeate the
membrane so as to dilute the solution.  A counter pressure
can be applied to the solution side to reverse water
transport.  The amount of pressure required to achieve a
static equilibrium is termed the osmotic pressure.  Reverse
osmosis is simply the application of a pressure greater than
the osmotic pressure which drives water from the solution
side of the membrane to the water side and permits
concentration of the solution.

The membrane plays the most critical role in the process.
By varying the properties of the membrane, one can control
the retention or passage of selected solutes.   When the
solute molecules are large; e.g., whey proteins, the osmotic
pressure is quite low and a membrane with relatively large
pores can be used at low operating pressures;  e.g., 10-50
psi .

-------
00
                          OSMOSIS
         I aim  I
                                   I otm.
                WATER
•SOLUTION


                       T
                    £  -U-  DILUTED
                    5  -i-*.  SOLUTION

                       T*"
                                          OSMOTIC
                                         'PRESSURE
                                      I I arm.
         WATER  PASSES THROUGH MEMBRANE  TO
         CAUSE  DILUTION  OF SOLUTION UNTIL OSMOTIC
         EQUILIBRATION IS ACHIEVED
                                REVERSE OSMOSIS OR UlTRAFilTRATiON
                                    I atm
                                                                    WATER
PRESSURE
>OSMOTIC
PRESSURE Of
SOLUTION
                                    PRESSURE IS USED TO DRIVE  WATER FROM
                                    SOLUTION

                                    IN MANY CASES SELECTED SOLUTES ARE
                                    ALSO DRIVEN FROM  THE SOLUTION
                             FIGURE 1,   OSMOSIS AND  REVERSE OSMOSIS

-------
This is termed ultrafiltration.  Referring to Figure 2, the
membrane may be called "louse"; i.e., lower molecular weight
solutes will pass through the membrane and will not be
retained in the concentrate.

In reverse osmosis, solutions of small molecules with
moderate to high osmotic pressures are retained and the
required driving force is considerably higher, ranging from
a few hundred psi to 800 psi, maximum in this system.
Higher operating pressures are required because of the
substantial osmotic pressure of salt and sugar solutions and
also because of the greater resistance to water transport of
RO membranes.  Again referring to Figure 2, a "tight"
membrane is employed.

Figure 3 shows a simplified flow sheet for the two-step whey
treatment process.  Cottage cheese whey, without filtration
for fines removal, is introduced into a low pressure UF unit
(step  1)'.  In this operation, whey is concentrated 12-fold
by volume.  Ultrafiltration membranes are used which retain
only the whey proteins.  Thus, it is possible to obtain a
protein concentrate with a higher proportion of proteins in
the disolved solids since lactose, non-protein nitrogen,
lactic acid, and minerals pass through the membrane.
Operation is typically in the pressure range of 15-50 psi,
and at temperatures of 120-125°F.  As was demonstrated, the
protein content of raw whey is increased from an initial
value of about 0.6% up to levels approaching 8% in this
step.  (In isolated cases protein content can be increased
to about 20$ but flux rates are so low as to become
impractical for this plant.)

The permeate (ultrafiltrate) from the UF unit is introduced
into a second membrane step.  In an RO operation, this
stream is concentrated from approximately 6% solids to
20-25$ solids.  Typical operation would be in the pressure
range, 500 to 800 psi, and at a temperature of 90°F.  The
membrane in the RO section is chosen so as to retain as
great a proportion of the organic solutes as possible,
resulting in the permeate having a lower BOD from 40,000
mg/1 to about 2000 mg/1.

The protein concentrate can either be used directly by
incorporation into food products or it can be dried.
Protein drying may be preceded by further concentration of
the retentate through conventional vacuum evaporation
equipment.  The lactose concentrate can be further
concentrated to about 50% solids by evaporation, and the
lactose can be recovered in a crystallization operation.

-------
H~*" *LOOSE*
SKIN
~.3,
>20A'PORES
             POROUS
             SUPPORT

SKIN
~.3MTHICK
~4A- PORES
POROUS
SUPPORT
     FIGURE 2,  MECHANISM OF MEMBRANE SEPARATI
        ON

-------
                       PROTEIN
                       CONCENTRATE
WHEY
STEP 1
                            UF
                                 10-50 PSIG
LACTOSE
                       H20,  LACTOSE
                       NON-PROTEIN  N,
                       LACTIC ACID
                       SALTS
                                                    CONCENTRATE
                                          STEP 2
                                                       RO
                                                    500-800  PSIG
                                                    LOW BOD
                                                    WATER
         FIGURE 3,  MEMBRANE PROCESS  FOR  WHEY  TREATMENT—FLOW SCHEMATIC

-------
Extensive prototype experiments were performed prior to the
design, installation, and operation of the pilot plant.
These were conducted at Abcor,  Inc. from November 1969
through June 1970, under subcontract to Crowley Foods, Inc,
In the course of these experiments, a variety of membranes
and membrane equipment were examined for both UF and RO.
Significant experimental results were presented and
discussed in the Phase I document, Water Pollution Control
Research Series 12060 DXF 07/71.
                            12

-------
                          SECTION V

         PHASE II:  300,000-POUND-PER-DAY WHEY PLANT


Figure 4 describes the industrial plant that is now
operational in LaFargeville.  In the membrane separation
processes for whey, the membranes used are thin films or
organic materials.

Because membranes are thin and have little strength, they
are always supported.  In our plant the membranes are inside
porous tubes which are held against supports by pressure.  A
porous tube with its membrane is referred to as a membrane
tube.  Two types of filters are used in the whey processing
plant - ultrafiltration membrane tubes and reverse osmosis
membrane 'tubes .

Ultrafiltration membrane tubes in this plant operate at a
maximum of 50 pounds per square inch gage (psig).  They are
connected together with plastic piping, and require no
additional fittings.  Reverse osmosis membrane tubes have
smaller pore size and operate at 800 psig.  These films are
inserted firmly into stainless steel castings.  Twenty
reverse osmosis membrane tubes are assembled as a group in
stainless steel connectors, and the nest is referred to as a
membrane unit.  Blocks of membranes have been combined in
sections to form building units.

An Abcor UF-480S module (an ultrafiltration module with 480
square feet of membranes) is a stainless steel cabinet
containing 216 ultrafiltration membrane tubes with all the
necessary piping and cleaning sprays.  The RO-850S (a
reverse osmosis module with 850 square feet of membranes) is
a stainless steel cabinet containing 63 reverse osmosis
units together with all the necessary piping and cleaning
sprays.  In this plant there are six UF-480S modules and six
RO-850S modules .

Ultrafiltration membranes are designed to retain.only large
molecules and to permit water and small molecules to pass
through them.  The solution which does not pass through the
membrane is referred to as concentrate.  During a batch run
of 150,000 pounds of whey, 137,500 pounds (11/12ths of the
whey) will pass through the ultrafiltration membrane.
Protein is accumulated in the concentrate and at the end of
the run will be present at about 1.2 percent.  Lactose,
lactic acid, and salts pass through the membrane.

                             13

-------
             UF
(SYSTEM CONTENTS  -  6,500  LBS)
  RO
                r£
                -H.
 CIRCULATION
 1,450 GPM
 FEED
180 GPM
(7) _
/±1
    CONCENTRATE
    12,500 LB
    AFTER 10 HRS
RETURN -  130-160  GPM
       SILO
       150,000 LB INITIAL
       6,000 LB FINAL
                                                                  - ^
                                 PERMEATE
                                 50-20 GPM
                                 137,500 LB/HR

                                 INTERSTAGE
                                  STORE
                 CONCENTRATE
                 6,9  GPM
                 3,MO  LB/HR
                                                                 PERMEATE
                                                                 20.7 GPM
                                                                 10,310 LB/HR
FEED
27.6 GPM
13,750 LB/HR
                     FIGURE 4.   WHEY PLANT SIMPLIFIED FLOW PLAN

-------
Thus, at the end of the run they are present in both the
concentrate and in the permeate at approximately their
original concentration in the whey.  Figure 5 shows a broad
view of the UF-RO plant, and Figure 6 shows two views of the
UF modules.

Ultrafiltration permeate feeds the reverse osmosis section.
Reverse osmosis membranes are supposed to retain all
molecules and to pass only water.  In fact they retain more
than 97% of the salts and about 95% of the lactose and
lactic acid molecules.  In this section of the plant
three-quarters of the feed will permeate.  The concentrate
leaving the plant will contain about H% salts and between 16
to 2Q% lactose and lactic acid.  The permeate discharge
contains about 0.07? salt with small amounts of lactose and
lactic acid which shows up as a positive BOD in a 1500-2500
mg/1 range.  The quantities of dissolved solids in the
permeate is a fixed fraction of the dissolved solids in the
concentrate.  Thus the more concentrated the lactose stream,
the higher will solids appear in the permeate.

The rate at which solution permeates membranes is called the
flux or flux rate.  Flux is usually reported in gallons per
square foot per 24-hour day (gfd).  (To use these units in
the ultrafiltration section, remember that each module
contains 480 square feet so that a flux of 1 gfd is
equivalent to a permeate flow of 0.333 gallon per minute
(gpm).  In a reverse osmosis module, which contains 850
square feet, a flux of 1 gfd produces a permeate flow of
0.591 gpm).

THE ULTRAFILTRATION (UF) PROCESS

QEerating_Cgnditions

In order to obtain the best continuous flux rate, certain
conditions are desirable.  Operating temperature should be
as high as possible without degrading or hurting the
membranes, which in ultrafiltration is between 120 and
125°F.  UF pressure should be held at about 50 psig.  Whey
should be pumped with Reynolds numbers of 8 - 12,000 inside
the membrane tubes to avoid film deposits.  Because flux
rate decreases when the concentration increases, the protein
concentration is kept as low as possible for the longest
possible period of time in a batch run.

UF whey is kept moving by two large circulation pumps in the
plant; one for standby and one for operation.  These pumps
delive between 1,200 and 1,450 gpm.  The rapid flow of whey
through the membrane tubes demands a large drop in pressure.
                               15

-------
-
             Figure 5.   Overai:  view of 300,000 pound-per-day UF/RO Whey Processing Plant during
                        installation at Crowley Foods,  Inc.,  La Fargeville,  New York  April 1972.

-------
Figure 6.  Views of Ultrafiltration Section and Complete Control Panel for
           300,000 pound-per-day Whey 1JF/RO Plant at Crowley Foods,  Inc.
           LaFargeville, New York  June 1972
',
                                     17

-------
Therefore , to maintain flux, pressure in this case is set
nut to go below 15 psig.  To protect the membrane tubes, the
pressure is regulated so as not to exceed 50 psig.  An
allowable pressure drop approximates 35 psig and is obtained
when whey flows through 18 membrane tubes in series via
manifolds through 12 parallel sets.  (Note FigXire 7-)
Contrul_uf_Concentration

In urder to achieve maximum flux rates over the system, one
needs to maintain the protein concentration as low as
possible in the circulating loop inside the membrane tubes.
As liquid permeates, more liquid is supplied to the
circulating loop in order to maintain the pressure.  There
are two feed pumps in the plant, but only one is normally
used at a time, while the other is being cleared or is
available for standby operation.  The pump feeds whey from
the storage silo into the circulating loop to make up for
the permeation.  However, the total permeation rate of whey
from this part of the plant is between 20 and 50 gpm, and
the feed pump is designed to supply 180 gpm.  The surplus
liquid from the feed pump leaves the circulating loop
through a valve and is returned to the storage silo.  The
minimum pressure in the circulating loop (15 psig) is
maintained by the pressure drop of the liquid through a
valve and is controlled by throttling this valve.  The high
rate of flow back to the storage silo insures that both the
contents of the silo and the contents of the circulating
loop become concentrated at the same time.  If some liquid
was not returned to the storage silo, the concentration in
the circulating loop would rapidly become high and for the
remainder of the run we would be permeating from a high
concentration.  By mixing back into the silo, we delay until
the end of the run, the time when the concentration reaches
its maximum.  The operation of the ultrafiltration section
is a batch operation.  The silo is filled with 150,000
pounds of whey, the circulating loop is filled, and
circulation is started.  As permeation proceeds, the
concentration in both the silo and in the circulating loop
increases and the level of liquid in the silo falls.  The
plant is designed so that at the end of ten hours 137,500
pounds of liquid will have permeated and 12,500 pounds of
concentrate will remain inside the circulating loop and in
the silo.   The level in the silo will be between one and two
feet and contain about 6,000 pounds of concentrate at the
end of the run.  The run is continued until the
concentration in the concentrate is as required; i.e., until
the level  in the silo drops to the predetermined point.
When the run is over, the concentrate must be recovered both
                             18

-------
Figure 7.  End view of Sanitary Whey Ultrafiltration Cabinet
           and Feed Circulation Pump  April 1972

-------
from the silo and from inside the membranes and piping and
pumped to concentrate storage.  When the first run is over,
the inside of the membranes will be quickly cleaned, and a
second run will begin.  When the concentrate from the second
run is recovered, the plant should be thoroughly cleaned and
sanitized before starting again.  The plant is able to
concentrate two batches, each having 150,000 pounds of whey,
in 20 hours, leaving four hours for cleaning up.


THE REVERSE OSMOSIS PROCESS (RO) (See Figure 8)

Like the ultrafiltration, this section of the whey plant
contains six modules so arranged that the feed solution,
which is the UF permeate, is fed to the modules in parallel.
Unlike a batch operation, the solution passes once through
the modules and its concentrate discharges continuously.
Each module contains 63 RO units, each having 13-5 square
feet, building the module to 850 square feet.  The 63 units
are arranged in 21 sets of three units through which liquid
flows in series.  Manifolds distribute the feed solution so
that the flow path is:  9 parallel sets, followed by 5
parallel sets, followed by 4 parallel sets, followed by 3
parallel sets.  As solution permeates out, the number of
parallel flow paths is decreased so that the linear velocity
of the concentrate inside the membrane tubes is maintained.

The RO membranes in this plant are supposed to hold back all
solute molecules and to pass only water.  The lactose
concentrate exhibits high osmotic pressure (approaching 400
psi for the exit); thus pressures this high must be used
inside the membrane tubes, both to overcome membrane
resistance.  Permeation rate is used approximately linearly
as the primary flow control in this section.  However,
during operation the pressure must not drop below the
osmotic pressure.  Pressure above 800 psi is avoided because
of possible damage to these membranes.

The high pressures needed for RO processing is provided by
sanitary homogenizer pumps.  There are two, but one is for
standby.  The plant is designed to process 13,750 pounds per
hour of feed (27.6 gpm), of which three-fourths permeates
(20.7 gpm) and one-fourth leaves as concentrate (6.9 gpm).
(Refer to Figure 9 - note the flow meters.)

Controlling_the_Flow_Rate

The fixed flow capacity of the pumps is 36 gpm.  This is
more than the design feed rate, so a small flow of solution
is at all times bypassed around the pump through a
                            20

-------
Figure 8.  View of Reverse Osmosis Section  of 300,000 pound-per-day  (140,000 liter-per-day)
           UF/RO Whey Processing Plant at Crowley Foods, Inc., LaFargeville, NY  April 1972

-------
          Figure 9.  View of Reverse
          Osmosis Pump and Interstate
          Tank of 300,000 pound-per-'
          d:Hv nlVRO Whey Processing
          Plant  April "]
-------
throttling valve.  RO feed rate is visible and can be read
on a flowmeter mounted on a panel next  to the pumps; the
feed rate is valve controlled.  If one  of the modules is not
in use, then the adjusting valve can be opened to lower the
feed rate to the system.  Concentrate  flow rate can be read
on its own flowmeter.  To obtain the design concentration,
the pressure is adjusted so that the concentra-te flow rate
is one-quarter of  the feed flow rate.   The pressure is
controlled by letting the concentrate  leaving the system
flow through a spring-loaded  back-pressure Regulator.  This
regulator must be  adjusted to obtain the design
concentration, provided only  that the  pressure is not
allowed to exceed  the limits  already mentioned.  Solution
flow through the RO  units shows a pressure drop of about 200
psi so the inlet and outlet pressures  are not the same.
                               23

-------
                         SECTION VI

                        PLANT DESIGN


DESIGN BASIS FOR ULTRAFILTRATION SECTION
The following design bases have been used for the UF
section:

1.   Twelve-fold volumetric concentration .

2.   Twenty hours operation per day in 2 batches; four
    hours for cleanup and sanitizing.

3.   Operating conditions:

          20 gpm per tube (1 inch id)
          120-125°F
          50 psi inlet and 15 psi outlet pressure to modules

4.   Average flux from pilot data was 13.8; published in
    Phase I Report 12060 DXF 07/71.

5.   Membrane area used:  300,000 Ibs./day:

       [Ibs                               [fractional
iQQiQQQ_day.l  x  24_£hrs/day.l  x  H/12_H2Q_remgval= 2810ft2
  8.5 [Ibs       20 [operating      1.38 [gal
       gal]          hrs/day]             day-ft2]
                            24

-------
A flow schematic for the ultrafiltration section is shown in
Figure 10.  The system contains  6 Abcor 480 ft2 modules,
providing a total membrane area  of 2880 ft2, slightly
exceeding the required 2810 ft2.  The six modules are
connected in a parallel flow arrangement.  The overall flow
design is analogous to that used in the pilot plant.

Operation is on a batch basis.   Whey is fed to a circulation
loop from a storage reservoir—insuring mixing between the
membrane loop and the storage reservoir.  This maximizes
both system capacity and protein yield.  At the end of a
batch operation, most of the protein concentrate is in the
silo, and tube content residual  is recovered from the system
by plug-flow flushing with water.  Ultrafiltration permeate
is withdrawn continuously and transferred to an interstage
tank prior to treatment by RO.

Raw whey is fed to the circulation loop with a 180 gpm feed
pump.  A second (spare) feed pump is included in the system
on continuous standby.  A high circulation rate through the
modules is maintained by a 60 HP centrifugal pump.  A second
(spare) circulation pump is included in the system, also on
continuous standby.

The ultrafiltration system contains a heat exchanger which
allows for either heating or cooling of the raw whey.  This
permits control over operating temperature in the system,
Additional instrumentation includes:

          - high/low pressure shutdown switch.
          - conductivity guards  on pumps.
          - sight glass on each  module to indicate flow.
          - temperature recorder/probe, to measure,
            control, and record  temperature.
          - pressure gauges to measure inlet and outlet
            operating pressures  (transmitted to control
            panel) .
                              25

-------
PERMEATE & CIP LINES
NOT SHOWN
i
I
1
J PIS
r^-fl .. .. •>_ DTTIIDM 1 T MP
    CONCENTRATE
      OUTLET   -
LEGEND
TRC - TEMPERATURE RECORDER CONTROLLER
  V - VENT VALVE
 HE - HEAT EXCHANGER
PIS - PRESSURE INDICATOR SWITCH
                                                               ONE  PUMP  IS
                                                                 STANDBY
                         SYSTEM
                         INLET
HE
(COTTAGE
 CHEESE WHEY)
      FIGURE 10,   SIMPLIFIED FLOW DIAGRAM OF ULTRAFILTRATION SYSTEM

-------
          - process valves, including two  to each module
            (on/off valves), controlled from operating
            panel.

Each module is fitted with CIP sprays and  piping,- and the
system includes a 250 gpm spray pump, piping, and valves.


DESIGN BASIS FOR  REVERSE OSMOSIS SECTION

1.   Four-fold volumetric concentration of  UF permeate to
    2\% solids.

2.   Twenty hours  of operation per day on a continuous once-
    through basis.  This is followed by a  four-hour cleanup
    and sanitizing.

3.   Operating conditions:  90°F, 800 psi.

4.   Average flux  of 5.85 gfd from pilot data.

5.   Membrane area requirements for 300,000 Ibs of whey per
    day are :

          x_.CH/121_x_L0^15_l  x  2i  x  (__i_) = 5100 ft 2
           (8.5)                 20      5.85
DESCRIPTION - REVERSE OSMOSIS SECTION

A flow schematic of the RO section is shown in Figure 11.
The system has 6 parallel passes, each containing 63 Abcor
RO 13-5 modules.  These modules have tubulence promoters and
AS-197 membranes.  Each parallel pass contains 850 ft 2
membrane area, corresponding to a total plant area of 5100
ft2.   The design for each pass is a once-through operation
with modules in a "Christmas-tree" arrangement.  Several
modules are connected in parallel at the beginning of each
pass, decreasing to a few modules at the end of each pass.

Each pass has its own valving arrangement so that it can be
shut down to replace any module which fails.  This piping
allows personnel to replace membranes with a shutdown of the
entire plant.  However, isolation of any single section of
the plant will lead to a temporary reduction of capacity of
M%.   Each pass has its source of filtered line water fed to
the pass through check valves permitting storage under line
water pressure.
                             27

-------
N>
00
                LEfifttfi

                CS -  CONDUCTIVITY  SWITCH
                HE -  HEAT  EXCHANGER
                TRC - TEMPERATURE  RECORDER CONTROLLER
                ACCUM -  ACCUMULATOR
                PI -  PRESSURE  INDICATOR
                SRV - SAFETY RELIEF VALVE
                FI -  FLOW  INDICATOR
                PS -  PRESSURE  SWITCH
                BPR - BACK PRESSURE REGULATOR
                  6000
                  GAL
                  TANK
     UF
PERMEATE  FI
                                                                                               LACTOSE
                                                                                             CONCENTRATE
                                      PUMP
                                     (STANDBY)
                                             FILTERED
                                              WATER
                                                                            PERMEATE FOR REUSE
                                                                            OR TO DRAIN
                 FIGURE  11.   SIMPLIFIED  FLOW DIAGRAM  OF  REVERSE OSMOSIS SYSTEM

-------
The membrane modules are fed by a  30 hp positive
displacement sanitary pump.  A s'-econd pump acts as a spare
in order to avoid plant shutdown in case of pump repairs.

Ultrafiltration permeate is collected in a 6,000 gallon
interstage tank (shown in Fig. 9)  and pumped through the
membrane modules on a once-through basis.  Its temperature
is controlled from the UF operating temperature to that of
the RO section by passage through  a heat exchanger (HE).

Permeate from the high pressure modules is manifolded and
discharged by gravity.  The plant  provided a reservoir for
this permeate with hope of reusing it for cheese processing
or cleaning.  However, it has not  been possible to
consistently control microbial growth on the outside of the
RO membranes.  Because of this, the microbial condition of
the permeate has not been suitable for re-use.  It has been
discharged to the sewer except for a period of several
weeks, in which it was treated with 8-10 ppm chlorine and
used for cleaning make-up water.
                            29

-------
                         SECTION VII

                       PLANT OPERATION
 The  physical plant that was constructed to house the
 equipment was able to receive machinery in early February
 1972.   By April  15, most of the membranes were in place and
 in various stages of erection.  Whey storage silos,
 electrical connections, and interstage piping still needed
 substantial amounts of connecting in this period.  Shortly
 after  this point in time the plant suddently seemed to take
 on the  appearance of a finished chemo-food processing
 installation.  A "shake down" or hydraulic testing period
 began  in mid May and the formal opening ceremonies were held
 un June 21, 1972.  A program of the occasion appears in the
 Appendix.' Figures 12, 13, and 14 show the whey plant
 attachment to the main cheese operations.

 During  the period May 16 to January 30, 1973, the whey plant
 was  operated about 125 days.  The other time periods of
 about  130 days were occupied in developing procedures for
 correcting the plant operational problems.

 Remarkable as it may seem we must note that the starting up
 of the  installation was straightforward and few serious
 place-on-line problems occurred; especially significant it
 seems  for a plant so large and complex.  Of course, we
 experienced the usual contractor delays that were
 distressing as were problems with some stainless steel bolts
 used on the RO module caps.

 Initially the ultrafiltration and reverse osmosis portions
 of the plant were able to process whey at rates above design
 standards.

 Data presented in Figures 15, 16, and  17 show that the plant
 was supplied with whey that was relatively uniform in pH,
 acidity, and total solids.  The cheese fines were not
 removed.  In about two weeks, however, the membrane system
 slowed in its capacity and its operation was unable to
 permeate flux rates according to plant design.

 We had not,  as yet, recognized a biological problem caused
 by improper clean-up in early June 1972 as evidenced by the
 Project Director's remarks made before the Whey Institute
meeting in Chicago in that same month.  A progress report
                             30

-------
  J
FIGURE 12,   FRONT VIEW OF CULTURED PRODUCTS PLAINT (WHEY BUILDING  LOCATED AT LOWER EXTREME
              LEFT), CROWLEY FOODS,  INC,,  UFARGEVILLE,  NEW YORK,

-------
....
' i
     FIGURE 13,  REAR VIEW OF CULTURED PRODUCTS PLANT  (WHEY BUILDING LOCATED AT LOWER RIGHT), CROWLEY FOODS,
                 INC,, LAFARGEVILLE, NEW YORK

-------

                                                                                                 •*«»
FIGURE 14.   ULTRAFILTRATION-REVERSE OSMOSIS BUILDING,  CROWLEY FOODS, INC,, I^ARGEVIIJLE, NEW YORK

-------
      4,5
LLJ
   SEPT
          NA   15
          X   4.57
         SD   0.05
      4.5
      4.0
                         n
                           »
                        1
 OCT
1972

 12
4.52
0.09
             JUNE


          N   22
          X  4.45
         SD  0.05

     AREFER TO GLOSSARY
                JULY
                1973

                 17
                4.45
                0.06

                                        NOV
                               11
                              4.60
                              0.18
              AUG


              21
             4.41
             0.06
FIGURE 15.   TYPICAL pH VALUES  OF WHEY SUPPLIED TO THE
           UF-RO PLANT:  HIGH, AVERAGE. AND Low VALUES
                      34

-------
-> CD
>- 0
1 	 rH
ca x
§ 2 50
LU <
OQ <_>
1 — i —
 CJ>
>- O
1 — •— 1
a x
§ 2 50
UJ <
II 1
CQ U
1 — t-
<3T O
ry D 2.74





















n,
"*^




* . *
OCT
1972
12
51.1
1.73

\
i '•;>
•)
j
JULY
1973
17
49.0
2.09





















n
i ;^
'•^m
'
.

NOV

11
50,3
2.28

H
' i
''r1!

AUG

22
49.0
1,69
FIGURE 16.  TYPICAL ACIDITY OF WHEY SUPPLIES TO THE
            UF-RO PLANT:  HIGH, AVERAGE, AND Low VALUES
                      35

-------
O
c/o
     6,00
     5,00
    SEPT


 N   12
 X  6.04
SD  0.21
O
CO
    6.00
    5.00
          X
         SD
             JUNE
 22
6.13
0.36
                             OCT
                            1972

                             12
                            6.24
                            0,16
                               _Lgj
                   JULY
                   1973

                    17
                   6.16
                   0.20
                               NOV
                               11
                              6.06
                              0.22
 AUG


 22
6.04
0.20
FIGURE 17,  TYPICAL TOTAL SOLIDS CONCENTRATION  IN  WHEY
            SUPPLIED TO THE UF-RO  PLANT:   HIGH,  AVERAGE,
            AND Low VALUES
                        36

-------
about the Crowley-EPA  Industrial  Grant  was  prepared for the
Whey Products Conference  and  is included  in  their
proceedings.

Unlike the  1970 Phase  I pilot  plant  work  in  Binghamton, New
York, the scaled-up whey  processing  plant presented some
special cleaning problems in  the  form of  8  inch tubing used
to collect module permeate  that had  been  underestimated.  In
addition, cleaning problems were  further  complicated early
in the plant's start-up when  operations required in feed
whey lines  from both the  new  and  old systems.  New lines
were physically connected to  old  nonsanitary whey collection
piping.  This cross connection hook-up  allowed lactobacillae
bacteria and others to contaminate the  whey  supply diverted
into the new installation.  Both  the 120°F  temperature and
the pH of fresh cottage cheese whey  produced from cottage
cheese manufacture provided an ideal environment for the
propagation of lactobacillae  organisms.

Inadvertently in the same period, to save time, we prepared
wash-up enzyme-cleaning solutions several hours before their
use needs.  This time  lag acted to dissipate cleaning power
of the enzyme system and  thereby  substantially reduced the
washing efficiency of  the plant's CIP procedures as
established for the whey  plant.

We soon discovered that 6 and  8-inch gravity permeate
collection  drain lines from the stainless steel modules into
surge tanks were not being  cleaned properly.  Where the
6-inch diameter piping was  used in pressure operations, we
found these areas were properly cleansed.   In the gravity
situations  the lines were not  filled completely with whey or
cleaning solutions and still  needed  an  alternate CIP spray
system or required manual cleaning operations.  In addition,
it must be  admitted that  the  dual equipment piping, which
was planned to provide processing flexibility, makes
clean-up even more difficult.  Standby  pumps with their
connecting  lines provide  difficult areas  to clean unless
appropriate cross connections  are properly  sanitized and
cleaned with other cleaning sections of the  installation in
the wash-up cycle.

ULTRAFILTRATION SECTION

Generally,  the UF section of  the  whey plant was found to be
mostly mechanically troublefree from June to December 1972.
Fifty tubes, about H%  of  the  system's total  tubing, were
replaced in the first  6 months of operation.  This kind of
experience  did not continue much  beyond January 19fj ana
will be explained later in  the text. Table  2 shows the
                             37

-------
early biological problems and correction by late August
1972. These improved cleaning results show as restored flux
rates in Figures 18 and 19-  These data suggest that the
system is, indeed,  cleanable and manageable when operated
and cleaned correctly.

Storing_Membranes

Because of bacteriological problems, it became necessary to
store the membrane banks of the large plant differently from
the Phase I project by continually circulating 5-10 ppm
chlorine sanitizing solution at 180 gpm when the plant was
idle.  The membrane modules were further exposed to an
exterior sanitizing spray while being sanitized internally
with circulating chlorine solution in its interior.  This
method was implemented by using the CIP circulating pump.
Additional sodium hypochlorite material was periodically
added at about three-hour intervals to the closed loop
circulation storage system to maintain a 6 ppm chlorine
residual.  In July when we apparently had not cleaned the
plant well, we found that chlorine dissipated, and it was
thought that its usage was a function of organic materials
left in the system after cleaning.  Later when we learned to
clean properly, it was shown that very little sanitizer had
to be added to the system when it was in the storage state.

In light of our early problems, cleaning cycles were changed
a number of times for trial periods in these first few
months.  Actually in the end, there was no real difference
in the final cleaning procedures for membranes or its
hardware beyond the original concept.  What was necessary,
however, was a need to develop more detailed directions for
operators undertactual operating conditions.  These specific
instructions given the operators are based upon the Abcor
Cleaning and Sanitizing Ultrafiltration and Reverse Osmosis
Systems, which is contained in the Appendix of this report.
The Crowley cleanup and storage procedure differs from the
Abcor recommended procedure in that pH 8.2-8.8 rather than
7-7-8.7 is used to maintain enzyme activity and that the
Quaternary Ammonium compound, Roccal, at 200 ppm is used to
store the UF unit rather than Zephiran recommended .

Tub_e_Rup_tures

Little experience had been obtained in the pilot plant on
how best to handle or replace malfunctioning UF tubes
because none had occurred in Binghamton in the Phase I part
of the project and only a few when the plant was moved to
LaFargeville for onsite use.  The actual time needed to
change a UF tube is small; less than 10 minutes, but the
                            38

-------
Table 2.  SELECTED MICROBIOLOGICAL DATA OF PRODUCTS PROCESSED IN THE ULTRAFILTRATION SECTION
                                           1972-73
Raw whey
Date
1972
5/16
6/2
6/7
6/23
6/28
7/4
7/5
7/19
7/28
8/1-8/30
9/1
9/11
9/15
9/18
9/19
10/14
10/18
10/23
10/31
SPCa Colib Yeast

_c _

-
Mold

-
Deproteinized whey
SPC Coli



500 <1
- <100 <1
100 10
9,000 10
1,000 10
<1,000 <10
<100 <10
<1,000 <10
<100 <10
- Modifications
<1,000 <10
600 <10
<1,000 <10
3,000 <10
7,000 <10
6,000 <10
320,000 <10
2,000 <10
1,000 <10
*
-
-
*
*
*
-
*
-
-
*
*
*
-
in cleaning
*
*
10
-
-
>100
>100
520
30
*
*
*
-
-
*
*
140
*

180,
840,
51,
3,
8,
800,
600 10
000 <10
000 <10
000 <10
000 <10
000 <10
000 <10
Yeast

*d
*
10
6
-
>100
>100
30
-
Mold

*
*
*
>100
-
>100
>100
*
-
UF
SPC


80,000
1

54
19
29
29
186
2
,600
100
,000
,000
,000
,000
,000
,000
concentrate
Coli Yeast

*^1 ^
<1 20
<10 *
<10 40
10 18
<10 >100
<10 >100
<10 >100
<10

Mold

*
70
*
*
*
>100
>100
*
-
methods
1,
4,
4,
4,

36,
5,
17,
1,
ooo <:LO
400 <10
ooo <:io
000 <10
900 <10
000 <10
000 <10
000
000 <10
*
960
640
-
-
>100
>100
-
60
*
*
48
-
-
100
100
-
*
<1
120
12
37
31
98
130

16
,000
,000
,000
,000
,000
,000
,000
_
,000
<10 *
<10 >100
<10 210
<10
<10
<10 80
<10 100
_ _
<10
*
>100
*
_
-
*
*
_
*
   d Standard Plate Count
   k Coliform
   c No information
   ^No count

-------
                 Table 2. (cont.)
SELECTED MICROBIOLOGICAL DATA OF PRODUCTS PROCESSED IN THE DLTRAFILTRATIOK
                           SECTION - 1972-73
4s.
O
Raw whey
Date
11/5
11/11
11/18
11/27
12/9
12/16
12/20
12/23
1973
1/10
1/19
1/22
1/30
Feb - No
3/7
3/14
3/21
3/28
4/4
4/11
4/18
4/25
5/2
5/9
SPC Coli Yeast
6,000 <10
3,000 <10
1,000 <10
9,600 <10
65,000 <10
9,000 <10
14,000 <10
6,000 <10

9,000 <10
30,000 <10
14,000 <10
2,000 <10
data generate -
2,000
1,000 <10
21,000 10
<1,000
-
<1,000
<1,000
<1,000
<1,000
2,000
40
*
*
*
100
*
10
*

90
>100
70
80
plant
*
*
70
60
-
*
*
10
*
*
Mold
*
*
*
*
*
*
*
*

*
100
*
*
under
*
*
*
*
-
*
*
*
*
*
Deproteinized whey
SPC Coli
1,
6,
4,

3,
2,
1,
7,

1,
3,
3,
1,
000 <10
000 40
000 <10
300 <10
000 <10
000 <10
000 <10
000 >10

000 <10
000 <10
000 <10
000 <10
repairs and not
<1,
<1>
6,
22,
372,
312,
1,300,
114,
44,
18,
000
000 <10
000 <10
000
000
000
000
000
000
000
Yeast
30
100
>100
>100
20
60
90
*

*
30
20
80
Mold
60
>100
*
*
*
80
*
90

*
*
*
*
UF
SPC
4,000
42,000
30,000
26,000
75,000
96,000
700,000
13,000

216,000
66,000
600,000
48,000
concentrate
Coli
<10
<10
<10
>100
<10
<10
<10
40

<10
<10
<10
<10
Yeast
*
100
*
100
30
>100
>100
30

30
>100
>100
>100
Mold
*
*
>100
90
60
30
*
*

40
>100
*
*
operating
*
*
60
*
*
70
390
60
580
20
*
*
*
*
*
*
*
*
510
*
<1,000
<1,000
<1,000
<1,000
3,000
<1,000
516,000
1,000
11,000
6,000
-
<10
<10
—
-
-
-
-
-
-
10
*
*
*
*
*
4,000
60
180
10
*
*
*
*
*
*
*
*
*
*

-------
Table 2. (cont.)
SELECTED MICROBIOLOGICAL DATA OF PRODUCTS PROCESSED IN THE ULTRAFILTRATION
                           SECTION - 1972-73
Raw whey
Date
5/16
5/23
5/30
6/6
6/13
6/20
6/27
7/4
7/11
7/18
7/25
8/1
8/8
8/15
8/22
8/29
9/5
9/12
9/19
9/26
10/3
10/10
10/17
10/24
10/31
SPC Coli
<1,000
1,000
<1,000
<1,000 <10
<1,000 <10
<1,000 <10
8,000 <10
<1,000 <10
1,000 <10
3,000 <10
<1,000 <10
<1,000 <10
1,000 <10
2,000 <10
1,000 <10
<1,000 <10
<1,000
<1,000
<1,000 <10
<1,000 <10
<1,000 <10
<1,000
<1,000 <10
<1,000 <10
<1,000 <10
Yeast Mold
*
*
*
140
*
*
*
190
30
300
*
30
40
1,400
*
20
560
*
10
*
*
*
*
*
*
*
*
*
*
*
*
10
*
*
10
*
20
*
300
*
*
*
90
*
*
*
A
A
*
*
Deproteinized whey
SPC Coli
62,
116,
96,
168,
113,
288,
132,
1,
310,
288,
840,
7,
700,
300,
370,
1,500,
43,
390,
9,
31,
156,
7,
<1,
168,
<1,
000
000
000
000 <10
000 180
000 80
000 <10
000 <10
000 <10
000 <10
000 <10
800
000
000
000
000
000
000
000 <10
000
000
000
000
000
000 <10
Yeast
30
10
10
180
190
604
60
10
20
30
360
20
*
760
1,840
3,100
1,880
980
*
280
4,300
1,180
*
720
*
Mold
*
*
*
A
10
*
10
*
10
*
20
*
A
*
A
*
*
*
*
20
300
*
A
*
*
UF concentrate
SPC Coli
2,000
1,000
9,000
1,100,000 <10
Yeast
90
10
50
50
Mold
60
A
10
50
48,000 <10 * *
6,000 <10

2,
4,
120,
2,
540,
264,
540,
504,
39,

570,
7,
1,000,
174,
960,
450,
82,
46,
300 <10
000 <10
000 <10
000
000 10
000
000
000
000
000
— _
000
000 <10
000
000
000
000
000
000 <10
46
60
30
20
290
50
380
960
400
620
110
_
120
20
30
A
20

A
*
A
20
A
A
4.0
A
*
A
$0
A
A
_
A
A
A
A
A
A
10
A

-------
Table 2. (cont.)
SELECTED MICROBIOLOGICAL DATA OF PRODUCTS PROCESSED IN THE ULTRAFILTRATION
                           SECTION - 1972-73
Raw whey
Date
11/7
11/14
11/21
11/28
12/5
12/12
12/19
SPC Coll
2
2
<1

<1
<1
<1
,000 <10
,000 <10
,000 <10
-
,000 <10
,000 <10
,000
Yeast
30
*
*
-
*
*
40
Mold
*
*
*
-
*
*
10
Deprotelnized whey
SPC
560,
330,
14,

6,
3,
1,
000
000
000
-
000
000
000
UF concentrate
Coll Yeast Mold SPC Coli Yeast Mold
4,000
2,500
<10 *
-
<10 50
<10 *
<10 *
90 500,
* 390,
* 18,
-
* 11,
* 340,
* 29,
000
000
000
-
000 <10
000
000
30
20
10
-
40
*
*
*
10
*
-
*
*
*

-------
                        PLANT   UNDER  CONSTRUCTION
             JAN
              FEB
 MAR
 APR
 X
 a:
    20
   15
 co
 _i
 ; 10
X
± 10
             SEPT
       X=17.34
      SD=3.33
               OCT            NOV           DEC
                        	—DESIGN FLUX
 FIGURE 18.  ULTRAFILTRATION —  PERMEATE FLUX DURING 1972:
             HIGH, AVERAGE, AND  Low VALUES FOR MONTH PERIODS
                               43

-------
I-A


 CD
 X


 cc
   20



   15

 _i

x 10
             MAY
                          JUNE
JULY
AUG
 I

 CD
 X


 o:
    20
 15  15
 _l


 >^  10
            SEPT
                           OCT             NOV           DEC
                                     ---- DESIGN FLUX
    FIGURE 19.   ULTRAFILTRATION  — PERMEATE FLUX DURING 1973:

                HIGH,  AVERAGE/ AND Low VALUES FOR MONTH PERIODS
                                 44

-------
effects of leaks and  ruptures  are  more  insidious.   Two major
problems occur when whey  protein  concentrate leaks  into  the
exterior system of UF membrane banks.   These materials
contaminate the deproteinized  liquids which feed  the  RO
portion of the whey plant.   It was previously reported that
whole whey with proteins  as  a  feed solution for RO  systems
makes them difficult  to clean. Normally bacteria cannot
pass through the UF membranes  when they are whole but they
can contaminate RO feed when they  escape from their closed
system.  In RO feed they  are troublesome.   Likewise it is
economically important that  the loss of proteins represents
a loss of approximately one  dollar per  pound of solids.
These losses can be high  if  concentrated protein streams are
lost near the end of  the  batch run.  Obviously the  time lost
to the operation, besides the  membrane  replacement  costs,
downgrade the economics of using  the system.

A positive system would be desirable to divert permeate from
UF modules back to the storage silos automatically when
problems occur; turbidity meters might  be  placed in line to
monitor UF permeate which is clear light green liquid when
proteins are removed.   This kind  of technology spin-off can
be normally expected  in equipment  evolution  modifications.

REVERSE OSMOSIS SECTION

This part of the plant was installed first.   Testing the RO
system for leaks and  pressure  began onsite  in mid-March
1972.  A closed loop  arrangement was connected by pumping
liquid from the interstage silo to the  banks  of membranes
and then back to the  silo.   Initially,  the  shipper's storing
solution was used for the wet  loop.  Once  committed to
moistening the system with solution, the plant then had to
use makeshift arrangements to  keep the  membranes
continuously wet to prevent  film drying  out  which would
spoil when dry.  Figure 20 shows the interstage silo and
others during this construction period.

Almost at once the RO membranes experienced malfunctions in
the stainless steel bolts holding  the units  stainless caps
in bundles.  This condition  allowed the  membrane's shipping
fluids to seep through their fiberglass  supports holding the
cellulose acetate membranes.   The  contractor  exchanged the
bolts for others with a different  type  of  steel-nickel alloy
of the same size but  these failed  too.   The  fact that the RO
banks contain more than 5000 bolts in the  system does point
out how over several  hundred hours of work  were expended in
changing different kinds  of  materials in an  attempt to
correct this problem; this problem was  a serious condition
until all the headers were drilled and  tapped to use 5/16
                            45

-------
Figure 20.  View of Whey Feed Silos and Interstage Tanks (with Portion of
            Two Reverse Osmosis Cabinets and RO Booster Pump Showing) at
            Whey Processing Plant, Crowley Foods, Inc., LaFargeville,
            New York  April 1972.
                                    46

-------
inch bolts rather than  the  1/4  inch  originally  supplied
The bolt problem was not  corrected until  mid-September  1972
By this time we had been  running  whey  products  through  the
RO system for more than  3 months.  Because  the-RO membranes
were subjected to leaking top and bottom  caps by bolt
shearing which allowed  whey  to  spray onto the exterior
portions of the modules,  yeast  and mold pro'blems were
encountered as shown in  Table 3.

The exterior CIP system  was  not designed  to  compensate for
gross product contaminating  conditions caused by bolt
breakage and therefore  was  inadequate  to  clean  the outsides
of the RO membranes supports in a satisfactory  manner.  The
CIP system appears to need more big  volume  spray nozzles
that can operate at elevated cleaning  pressure.

The plant personnel spent many  man-hours  of  daily difficult
manual work in the attempt to clean  the individual membrane
exteriors so they could  be free of molds  and yeast.  The
situation was worsened  by air contaminatants because the RO
modules were run without  top covers  during  this period as
there was the need to check  the membranes frequently for
leakage due to the bolt  problems.  The double trouble to the
plant was that both sections caused  problems simultaneously
and this situation made  the  entire project less than
satisfactory to manage.   We  would have probably been better
able to resolve the RO  difficulties  sooner had the UF
portion not caused problems  during this period.

Qoncep_t ^Differences

The RO membranes are arranged in a pyramid of
"Christmas-tree" flow pattern.  Our  previous experience with
Phase I in the pilot plant was  rather in a three section
circulating loop RO plant.   Each section in the small model
had its own pump forming  a circulation loop that maintained
fluids in a continuous  turbulent-like condition.  To obtain
this same turbulence, turbulence promoters were used in the
membranes in the full scale  plant.   These are plastic beads
7/16 inches in diameter.  Approximately 145 are placed
inside each tube.  The membranes for the big plant are
somewhat different in membrane  numbers per individual module
as well as being -secured  by  a newer  type seal with a plastic
collar versus epoxy glue  in  our previous work.  Turbulence
promoters were used in  the full scale plant compared to no
turbulence promoters in  the  earlier  model.  These
differences may account  for  some additional difficulWes^n
keeping the system as clean  as  the Phase I pilot plant with
its simple circulation  technique.
                             47

-------
                    Table 3.   TYPICAL MICROBIOLOGICAL  DATA OF  PRODUCTS PROCESSED IN THE REVERSE OSMOSIS SECTION
                                                              1972-73
oo
Deproteinized whey
Date
1972
5/16
6/3
6/6
6/7
7/5
8/2
9/1
9/11
9/19
10/16
10/18
10/23
10/31
11/6
11/20
11/25
11/27
TSa

5.46
6.03
-g
5.74
5.63
6.16
-
5.90
5.41
5.58
5.47
5.32
5.35
5.22
5.88
5.56
5.67
SPCb

500
<400
2,500
400
3,000
30,000
9,000
76,000
900
19,000
5,000
17,000
<1,000
3,000
10,200
1,000
300
Colic Yd

<1 *f
<1 *
20 <10
<10 >100
<40 >100
<10 40
<10 >100
<10 >100
<10
<10 >100
<10 >100
<10 *
<10 60
<10 >100
<10 100
<10 80
<10 110
Me

*
*
<10
<10
>100
*
>100
>100
-
>100
100
*
*
A
50
*
*
TS

13.61
16.98
19.97
18.71
17.9
18.54
19.38
19.28
18.97
19.72
18.86
18.76
16.37
17.13
19.34
16.70
19.49
RO concentrate
SPC Coli

2,600 <1
3,600 <1
318,000 40
500,000 100
tntc 61
40
700,000 <10
>1, 000, 000 30
600,000 <10
4,000 <10
500,000 <10
59,000 <10
16,000 <10
19,000 <10
24,000 <10
5,000 <10
7,700 <10
Y

*
*
>100
>100
>100
>100
>100
>100
-
>100
>100
>100
100
A
>100
>100
>100
M

*
A
>100
>100
>100
>100
>100
>100
-
*
A
A
A
*
10
100
*
RO
SPC

200
800
-
-
490,000
30,000
39,000
41,000
620,000
128,000
<1,000
51,000
3,000
13,000
300
2,000
3,500
permeate
Coli Y

100
<10
<10 >100
<10 >100
<10
10 100
<10 *
<10 >100
<10 >100
<10
<10 >100
<10 >100
<10 >100

M

A
A
-
-
>100
-
>100
>100
-
100
A
>100
A
-
<10
>100
*
                      , Total solids in percent
                       Standard Plate Count
                      . Coliform
                       Yeast
iMold
t No count
' No information

-------
                  Table 3.  (cont)
TYPICAL MICROBIOLOGICAL DATA OF PRODUCTS PROCESSED IN THE REVERSE OSMOSIS
                           SECTION - 1972-73
vo
Deproteinized whey
Date
12/2
12/6
12/16
12/20
12/27
1973
1/3
1/8
1/10
1/19
1/20
1/22
1/29
1/30
TS
5.41
5.18
5.55
-
5.60

5.43
5.4
5.51
5.40
5.69
4.73
5.58
5.59
SPC Coli
<1,000 <10
<1,000 <10
2,000 <10
1,000 <10
<1,000 <10

1,200 70
<1,000 <10
<1,000 <10
3,000 <10
2,000 <10
3,000 <10
5,000 <10
1,000 <10
Y
30
100
60
90
100

40
50
*
30
*
20
MOO
90
M
40
*
80
*
*

*
*
*
*
10
*
*
*
TS
19.22
15.69
18.40
17.84
17.99

16.14
12.95
14.09
16.17
15.64
16.32
17.53
18.55
3.0 concentrate
SPC
73,000
13,000
29,000
8,000
27 , 000

6,700
16,000
19,000
24,000
18,000
34,000
32,000
44,000
Coli
<10
60
70
<10
100

MOO
<10
<10
20
<10
<10
50
<10
Y M
_ _
MOO MOO
MOO MOO
60 100
MOO MOO

100 100
100 *
100 100
100 100
100 *
100 90
100 *
100 90
RO permeate
SPC
28,000
21,000
6,000
89,000
27,000

„
-
384,000
336,000
18,000
79,000
26,000
19,500
Coli Y
<10 70
<10 MOO
<10 MOO
<10 MOO
<10 MOO

_ _
- -
<10 MOO
10 MOO
<10 MOO
<10 MOO
<10 MOO
<10 MOO
M
>50
MOO
MOO
20
MOO

_
-
MOO
MOO
MOO
MOO
20
60

-------
Two RO operational problems are apparent in the Phase II
operations.  One situation is that product runs require one
set of flow rate values and these basically are now being
met.  The other area is that cleaning flow rates and volume
needs are different than those required in the product flow
configurations.  Opinions of the project's engineers
differed in how best to resolve the problem.  During
November and December 1972 much data were collected to
pinpoint the problems.  Corrective measures were taken to
install bypass CIP valving and plumbing that allow improved
cleaning in membrane tubes with a method that increases
volume, velocity, and turbulence.  A diagram showing these
changes is in the Appendix of this report.

In  Phase I we were able to keep the pilot plant equipment in
a desirable state of cleanliness with only water.  We fully
anticipated that we should have been able to repeat the
Phase I experience with the manufacturer's new
modifications.  In order to improve cleaning and rule out
water quality problems in this location which has hard and
high solids water, we installed water softening equipment to
improve cleanability water characteristics.  Information
about the softening equipment is supplied in the Appendix
with selected water analysis data.

FLUX RATES

Initially the RO section performed in excess of design
specifications.  Each of the 6 modules contained 63 RO
units.  The unit composed of tubes having a total capacity
of  13-5 ft^ area.  Quite early in June, in order to balance
the pressure drops and maintain the flow rates,  3 RO units
were removed from each module.  From that time in June until
January 1973, the total RO section did not run at its
specified capacity as shown in Figures 21 & 22.   Because
membrane fouling seemed obvious, a number of methods were
employed to clean the membranes.  Enzyme cleaners, as well
as  chlorinated materials, were pumped into the system and
these did help to keep the system clean.  Chemical cleaning
improved bacteriological results but this condition lasts
briefly and failed to bring the RO plant capacity back to
its start-up flux rates.

In January 1973, Abcor engineers finally recognized that
there was indeed a special kind of cleaning problem in the
RO section caused by a design or chemical problem.  Their
tests indicated to them that higher flow rate cleaning
cycles would improve cleaning operations.  By making flow
path changes in piping, the flow rate would be increased
during the CIP cleaning process.  Plumbing changes were made
                            50

-------
                        PLANT  UNDER   CONSTRUCTION
            JAN
FEE
MAR
APR
x
   21
   17
   13
              o-	
            MAY

JUNE
JULY
                                                           o-
AUG
 <
 o
   21
   17
   13
           SEPT


       N=82

       X=16.49
 OCT           NOV
                                                         DEC
         	 DESIGN FLUX
    FIGURE 21.   REVERSE OSMOSIS - PERMEATE  FLUX DURING 1972:

                HIGH,  AVERAGE, AND Low VALUES  FOR MONTH PERIODS
                                  51

-------
   23
	
                                                         •a.
            MAY
   JUNE
JULY
AUG
           SEPT
      SD=2.61
                                   	 DESIGN FLUX
    FIGURE  22,   REVERSE  OSMOSIS —  PERMEATE  FLUX DURING 1973:

                HIGH, AVERAGE, AND  Low VALUES  FOR MONTH PERIODS
                                 52

-------
                          PLANT  UNDER  CONSTRUCTION
              JAN
 FEE
                MAR
                                                            APR
I
CD
   3.0
 X
 O
 s:
CD
PQ
   2.0
            ce oo
              MAY
JUNE
               
               (=) UJ
                                           CD
                                                            
-------
   3.0
CD
C/D
   2,0
   1,0
             JAN
     H     o	,
 FEB
                                        MAR
APR
CX)
C=3
3.0

2,0

1.0
             MAY
                                  (=)
                                CD — I
                                2Z CD
                                      O

                                      o


                                      O
JUNE
                                        JULY
AUG
             SEPT
       N=188
       1=1.21
      SD=0.66

FIGURE 24,   RO PERMEATE QUALITY  DURING  1973  AS  MEASURED BY TOTAL
             SOLIDS WHERE 1% APPROXIMATES  7000 MG/L  BOD5:
            HIGH, AVERAGE, AND Low VALUES  FOR 10-DAY  PERIODS
                                54

-------
in flow  paths to modify  the cleaning  procedures   RO flux
rates  improved in January by replacing  the first 9 RO units
in each  module.  Abcor  reported that  the  first few RO units
in each  module had  lost  some flux capacity by their
examination  for unknown  reasons.  It  was  suspected that
cleaners and sanitizing  chemicals used  to clean-previous
fouling  conditions  caused changes in  membrane structure
When   these  few units are replaced, the change should
correct  the  RO plant and bring the total  units back to
capacity.   Figures  23 &  24 show typical BOD results in the
permeate generated  from  the systems,  and  Table 4 shows
selected microbial  problems in concentrate and permeate.

       Table  4.  SELECTED DATA TO ILLUSTRATE BACTERIOLOGICAL RESULTS IN RO
                          CONCENTRATE AND PERMEATE
Concentrate
Date
19^2
10/17
10/24
10/31
11/7
11/14
11/21


12/5
12/12
12/19
SPC Coli

1,500,000 -a
624,000 -
2,280,000 -
380,000 -
486,000 -
1,140,000 -
Spt

142,000 -
73,000 -
16,000 -
Yeast

28,000
3,600
>3,000
>3,000
>3,000
>3,000
Mold

10
120
>3,600
1,200
740
2,400
jcial Cleaning Qhemi
64,000
3,100
480
130
10
*
Permeate
SPC Coli

2,340,000 -
2,000,000 -
3,600,000 -
430,000 -
300,000 -
156,000 -
^als

25,000 -
12,000 -
726,000 -
Yeast

>3,ooo
>3,ooo
>3,000
7,200
2,400
>300


520
*
14,000
Mold

>3,ooo
>3,ooo
20
60
20
20


*b
*
90
  a No information
  b No count
Reverse Osmosis Unit Replacement

Eight units in total were replaced in the RO section because
of rupture or leaking during the  first 8 months that they
were in operation.  Eighteen units were removed in June  1972
to correct a pressure drop in the system.  One might look
upon this as a credit but this is not a valid appraisal.
The present situation seems to suggest that we need to
provide  9 units for  6 modules,  or a total  of  54 units are
needed to make the system perform at design  specification.
If we expressed these as a percentage  figure  applied to
membrane  replacement, this would confuse the  situation as we
view the  RO section  at  this  time.
                                55

-------
Mechanical_Prgblems

Work was still underway, as of March 1973, to correct the
mechanical problems that remain in the RO section of the
plant.  Some maladjustments still remain in flow rates and
potential plugging as well as the detection methods to find
leaks when they occur.

Bacterial numbers in RO concentrate and permeate increased
substantially between January-October 1973-  Crowley
personnel elected to resolve UF operational problems before
altering conditions that troubled the RO portion of the
plant.

Standard Plate Counts (SPC) in both fractions generally ran
in the 500,000 to 2 million/ml of product during the
aforementioned period.  Yeast and mold counts combined
fluctuated about in a one million range.

RO concentrate was wasted to land disposal areas, and RO
permeate was discharged as wastewater to the site's sewage
disposal plant.

Microbial counts were controlled in material used for
product by simply pasteurizing the liquids as they left the
RO system.

The RO portion of the plant was operated in this manner
until mid 197^ when the Crowley firm decided to re-membrane
the RO modules by stages.  Figure 25 shows data obtained
from the first cabinet reworked with new membranes.  Flux
an-d BOD levels performed according to design standards.
Because of the RO mechanical difficulties and sanitation
problems, continued efforts for correction are necessary by
Crowley and Abcor personnel.  These efforts and results will
be covered in future annual reports.
                              S6

-------
                                                      IW
                                                      X=20.71
                                                     SD=1.93
  cc
  X
 
-------
                        SECTION VIII

                      PROCESS ECONOMICS
 Preliminary economic data should be viewed with caution
 because they reflect high start-up cost, very low product
 yield over time, and too many hours of executive labor.
 Table 5 summarizes the capital cost .

 Table 5.  DOLLAR COSTS FOR SETTING UP A 300,000 LB/DAY
                         WHEY PLANT

 ______________________ Cost __________________________ Amount

     Construction contracts                   $613,655-76
     Equipment installation                     59,108.67
     Contingency equipment and contracts        18,421.83
     Firm's expenditures in salaries and        67,912.27
        overhead for project
 In addition to the above capital cost, $186,220.54, was
 spent on operation and maintenance during the start-up
 period.  This expense was caused by the need to correct
 technical problems, which should have been covered in the
 research development phase.

 Time spent in the as yet non-profitable whey operations have
 detracted executive time substantially from their other
 activities that might have enhanced the profitability of
 Crowley's.  Subsequent costs are in Appendix.

 Personnel on site other than whey plant people were also
 involved in day-to-day problems of ordering supplies or
monitoring contractors, etc. to get the plant operating and
 then provide support actions to keep it operational.  Much
of these costs are not reflected in the data that follow.

On the other hand, it seems unrealistic to charge expenses
against the whey plant that are "one-shot type cost" items.
Many of the expenditures are neither capital nor variable
expenses but better designated as some new technology charge
for marrying new processes into existing operations.
                             58

-------
The general format of reporting plant  expenses will  follow
the method used  in our  Phase I report.   This method  allows
one to  compare  the Phase  I  cost projections with actual
experience .

The expense of  operating  the whey plant  at LaFargeville,
which  is a 6-day,  24-hour per day operation, has been
computed for  5  months.  We  selected  the  period of August
through December because  in this period  we achieved
industrial plant-like operations.  The  Phase I report
figures are repeated along  with the  Phase II dollars but  in
5/12th amounts  to  account for 5 months  operation of  the
pilot  works annual projections as shown  in Table 6.

 TABLE 6.  FIVE MONTHS  PROJECTED3 AND  ACTUAL OPERATING COSTS
                  FOR 300,000 LB/DAY WHEY  PLANT
                                                     Difference
               Costs	Projection    Actual  Actual-Projection
     Labor  and overhead
       Supervisory          $  2,000      $12,242      $10,242
       General utility         9,500       7,921        1,079
       Laboratory                800       1,403         603
       Maintenance               535         607         132
       Fringe benefits         5,150       8,870        3,720
         and overhead
         (40$ of wages)                                     ,
       Membrane replacement    16,625       1,022b       15,603°
       Power                   2,100       3,941        1,841
       Water                   1,650          nac         na
       Steam                   1,250       1,560         319
       Cleaning suppliesd         800       5,996        5,196
       Professional services    4,165       4,165            0
       Depreciation           36,790      20,255        6,545
          aBased on Phase  I data             ^, ^ u      107-3
          bDoes not include membrane failures of February 1973
          cNo available
          dTable 7 describes cleaning chemicals
                                  59

-------
Table 7.  WHEY PLANT CLEANING SUPPLIES USAGE
August
Cleaning
product
Ultra-Clean
Enbiozyme
Antibac B
Caustic
soda
102 Liquid
chlorine
Alcozyme
Rhozyme
Salt beads
Total Cost
Amount
used
1080 Ibs
12 gals
350 Ibs

315 Ibs
60 gals
300 Ibs
50 Ibs
-

Cost
$ 810.
54.
232.

22.
30.
195.
41.
-
$1385.
September
Amount
used Cost
00 1400 Ibs $1050.00
00 37 gals 166.50
50 350 Ibs 232.50

84 300 Ibs 21.75
00 60 gals 30.00
00 200 Ibs 130.00
00
-
34 $1630.75
October November December
TOTAL
Amount Amount Amount
used Cost used Cost used Cost
817 Ibs $ 612.75 803 Ibs $602.25 600 Ibs $450
37 gals 166.50 24 gals 108.00 24 gals 108
350 Ibs 232.50 350 Ibs 232.50 350 Ibs 232

315 Ibs 22.84 180 Ibs 13.05 96 Ibs 7
60 gals 30.00 60 gals 30.00 20 gals 10
100 Ibs 65.00
_
3300 Ibs 57
$1129.59 $985.80 $964
.00
.00
.50

.00
.00
-
-
.42
.88
Amount
used
4700
134
1750

1206
260
600
50
3300

Ibs
gals
Ibs

Ibs
gals
Ibs
Ibs
Ibs

Cost
$3525.00
603.
1162.

87.
130.
390.
41.
57.
$5996,
00
50

44
.00
.00
.00
.42
.36

-------
These data fail  to  indicate  how much  product  was made nor
does it reflect  by-product  flow rates in  unit costs.  Plant
expenses, for the most  part,  would  not have been overly
different if more whey  rather than  water  was  pumped through
the membrane system.

Realistic actual unit  processing costs are not  possible in
this period because the installation  had  not  as yet
exploited its full  whey handling volume potential.  Future
annual reports will indicate better the real  impact of
operating costs  on  whey processing  when the system achieves
prolonged stability.
Dep_reciatign_and_Interest

Interest  costs  have been calculated at  8$.   This amount may
well  be understated with present inflation  and high rate
borrowing.   The total cost,  less the EPA reimbursement of
$360,000,  is used as Crowley's base cost for depreciation
and other  pertinent calculations.

More  recent data than these  generated at start up are listed
in Tables  8, 9, and 10,  which summarize operational costs of
one year  between May 1973-April 1974.  Membrane replacement
costs do  reflect conservative projections in view of initial
experiments with membrane failures.
                              61

-------
   TABLE 8.  OPERATING COST DURING MAY 1973 - APRIL  1974 a
     Supervisor's salary                          $  10,500
     Plant salaries and fring benefits               47,972
     Membranes                                       78,600
     Cleaning supplies                               27,679
     Electricity                                      9,084
     Fuel.                                            2,355
     Repair and maintenance                          15,345
     Property taxes, insurance, etc.                 19,512
     Administration                                   6,000
     Laboratory                                       4,500
     Total                 ,                       $ 221,547
     aDoes not include building and equipment depreciation
         membrane life - 8 months
      RO membrane life - 12 months

     csee Table 9
During the 12-month period, May 1973 - April 1974,
52,608,000 pounds of whey, containing 3,226,532 pounds of
solids, were produced.
                             62

-------
     Table 9-  PRODUCT  FLOW SUMMARY OF UF-RO  OPERATION
                    MAY  1973 TO APRIL 1974
                             Product
                                                  Pounds
Recovered

Protein in form of  concentrate
Other solids in the UF  concentrate3
Other solids condensed  to  15-20$ in RO

     Total whey solids

Not_Recovered

Solids in RO effluent
Unaccounted for solids
                                                  200,751
                                                  200,751
                                               1^89.8.138

                                                2,299,640
                                                  564,415
                                                  362,477
aProtein concentrate  is  estimated  to average 50$ protein on
dry basis.

      Table  10.  CLEANING  SUPPLIES FOR THE UF-RO PLANT,
                   MAY  1973  TO  APRIL 197^
Cleaning Product3
Ultra-Clean
Erabiozyme
Antibac B
Caustic soda
Liquid chlorine
Bonewitz C-4
Glow acid
Phos-acid
Roccal
Salt beads
Unit cost
$ 116.00/cwt
5.50/gal
68.40/cwt
12. 12/cwt
.65/gal
23-57/cwt
51 .31/cwt
22.71/cwt
4.22/gal
3. 12/cwt
Usage/day
42
4
3
6.25
4
25
10
1
1
100
Ib
gal
Ib
Ib
gal
Ib
Ib
gal
T
gal
Ib
                             63

-------
                         SECTION IX

              MEMBRANE REPLACEMENT OF UF TUBES
UF membrane failures occurred abnormally high in January
1973.  A total of  18 tubes had to be replaced in 15 runs.
Five tubes needed  replacement on January 20, 1973 which
until this episode was an all-time high for the number of
tube failures in one day.

On February 2, 1973 the plant operator detected additional
membrane failures  and replaced 10 UF tubes.  Still, the UF
permeate remained  somewhat turbid, suggesting that protein
seepage into UF permeate was occurring.  The boil testing of
whey failed to clean the solution which is an excellent
method to determine if whey proteins are escaping from their
membrane system.

By February 8, 1973, the situation was declared critical
enough to warrant  action to disassemble and inspect each
tube in cabinet number one Forty-four tubes of the total 216
in the module were found to be defective.  The plant
personnel found that defective tube conditions varied from
slight membrane bubbles to loose films and even some with
shredded membranes.

The tube failure situation occurred after a slowdown in UF
flux rates.  Plant personnel attempted to correct the system
for lack of capacity by using a sponge ball cleaning method
(see Appendix) that has been successful to correct low flux
in the past.  Many faulty tubes were discovered during this
procedure.  A decision was made that each of the 1296 UF
tubes be examined  for outward defects.  All tubes were
removed from the cabinets and returned to Abcor, Boston,
Massachusetts, for examination.  A chronological order of
failures leading to this decision appear in Appendix E.

TYPES OF FAILURE

1.   Some tubes had ruptured membranes.

2.   Many membranes appeared to have lost their elasticity or
    were found to be poorly elastic.

3.   Membranes in many cases lost their adhesive bond to
    their membrane carrier tube.
                             64

-------
4.  Many tubes had  lost  their opaque color.

Later in this investigation it was found that about  30% of
the tubes considered  having failed were  able  to  be brought
back to a normal  state  in  Abcor's Boston Laboratory  by
special cleaning.   However, generally,  it was felt that
about 70$ of the  total  UF  tubes were unusable.

REASONS FOR FAILURE

Paramount to the  bad  experience of having a massive  tube
failure throughout  the  UF  section of the plant was the need
to determine why  the  trouble occurred.   If one were  to
assume that the  life  of  the membranes had simply been
exhausted, the economic  effect upon the  process is obvious.
Other operations  in and  out of continental USA have  shown
that membranes have been operating more  than  2 years without
a mass failure.   Table  6 does not reflect this event for it
may not be typical  in this time frame.

Both Crowley scientists  and the Abcor team reviewed  in depth
the events preceding  the membrane failures.   In our  opinion
the following combined  situations created conditions that
caused UF membrane  failures.

1 .  Circulations  of interior and exterior UF  membranes in a
    6-10 ppm chlorine solution reduced the life expectancy
    of UF cellulose acetate membranes by de-acetylation .

2.  The hours of  sanitizing storage was  disproportionate by
    many fold the hours  of whey operation through the period
    of July to February, storing time was more than  running
    time .

3.  Caustic is added  to  enzyme cleaners  to maintain  a
    8.2-8.8 pH,  this  range hastens hydrolysis of delicate
    membranes.   Some  cleaning chemicals  may have not
    dissolved or  failed  to be uniformly  mixed into cleaning
    solutions thereby it seems reasonable that some  washing
    solutions were  used  at higher concentration than our
    records show.   This  failure could occur in a "slug
    type" fluid  feed  situation.

4   SoonKe balling  membranes to remove fouling conditions
                 the membrane bonding to  its backing  more
    than we  realize.
5.  Manufacturing  quality  control  tolerances of these
    membranes  and  their  supports may  have  to be st.ll
                             65

-------
    further improved.   Abcor has been producing new
    generations of more durable membranes with fiberglass
    backings.   As yet,  experimental data are not available
    to predict the characteristics of these new materials.

Appendix E lists UF and RO membrane defects as to date,
location, and  type of failure.   Little correlation appears
visible by these records to explain when or where failures
occur .
                           66

-------
                           SECTION  X

                   PERTINENT  PUBLICATIONS
1.   Goldsmith, R. L., D. J.  Goldstein, B. S. Horton,
    S. Houssain, and  R.  R.  Zall.   Ultrafiltration for
    Control of Pollution Caused by Cottage Cheese Whey.
    Industrie Aliment.   Pinerolo,  Italy.  October 1971.

2.   Zall, R.  R. Membrane Processing of 300,000 Pounds Per
    Day Cottage Cheese  Whey  for Pollution Abatement -
    Phase II.  AlChe  64th  Annual Meeting.  San Francisco,
    California.  November  1971.

3.   Horton, B. S.,  R. L. Goldsmith, and R. R. Zall.
    Membrane  Processing of  Cheese  Whey Reaches Commercial
    Scale.  Food Tech.  26:30-32, 34-35.  February 1972.

4.   Horton, B. S.,  R. L. Goldsmith, and R. R. Zall.
    Traitment sur Membrane  du  Serum de Fromage au Stade
    Industrie!.  Rev. Lait.  Francaise.  298:367-379.
    May  1972.

5.   Zall, R.  R. and D.  J.  Goldstein.  Membrane Processing
    of Cottage Cheese Whey.  Proceedings of the Whey
    Institute Conference.   Chicago, Illinois.  June 1972.

6.   Selitzer, R. Crowley Begins Membrane Processing of
    Cottage Cheese  Whey.   Dairy and Ice Cream Field
    155:34-41.  June  1972.
                            67

-------
                         SECTION XI

                          GLOSSARY


Angstromi.._A^ - 10^ centimeter.

Bactericide - Any material capable of inhibiting or
                destroying bacteria.

BOD - BOD5 always in this report.  Biochemical oxygen
        demand, an indirect measure of the concentration of
        biologically degradable material.

Brackish - Somewhat salty, but substantially less so than
             seawater .

CejLlulgse_acet.ate - A water insoluble polymer formed by the
                      reaction of certain acids on
                      cellulose.

Chlgrinatign - The addition of small amounts of free
                 chlorine, usually to water, for the purpose
                 of killing harmful organisms in the water
                 and rendering it safe for drinking.

COD - Chemical oxygen demand, an indirect measure of the
        concentration of organic matter oxidized by a
        boiling acid digestion system.  COD values can be
        correlated with BOD values from the same wastes.

Effluent - The residual output stream, usually the waste.

Filtrate - That which has passed through the membrane,
             usually the solvent.

Flux - The membrane throughput, usually expressed as some
         volume per unit time.  See:  gfd, gpd.

Fouling - The act depositing on the membrane surface some-
            thing which will impede it proper functioning,
            sometimes also termed "blinding",

gfd - Gallons per square foot per day - gal/ft^/day.
                             68

-------
  d - Gallons per  day.

Membrane - The thin  anisotropic  polymer film,  a  substance
             with  different properties in  different
             direction,  which is the  active  separation
             mechanism  in osmosis,  reverse osmosis, and
             ultraf iltration .

Module - The combination of a membrane element and its
           pressure  container .

N - Number of observations .

Osmosis - Self-diffusion through a  semipermeable membrane of
            a solvent due to the differential pressure
            between  two solutions of  differing
            concentrations .

Permeable - Capable  of  allowing  some  material to pass
              through essentially unimpeded.

Permeate - That  which passes through.

p_H - The negative  logarithm of the  hydrogen  ion
       concentration, scaled 0-14,  with 0-7  representing
       relative  acidity, 7 being neutral,  and 7 through
       14 relative alkalinity.

Pore - An imagined opening in a  membrane which allows
         certain components, but not  others, to pass
         through .

Porous - Having  physical strength and  appearing solid but
           capable of allowing certain materials to pass
           through virtually unimpeded.

psU Pressure -  (1)  "Pounds per  square inch."  (2) The
                   normalized units  of  force.

Rejection - The  amount  of material  not allowed to pass
              through a membrane .
                     sufficient reverse  pressure to cause the
                     solvent to flow in  its  unnatural
                     direction .
SD  -  Standard  deviation
                              69

-------
Semi - Technically means "half" but more usually used to
         mean "partially."

Ultrafiltratign - An ultrafine filtering action generally
                    applied to solutes with a molecular
                    weight above several hundred and capable
                    of being strained by pores in the region
                    of a few hundred Angstroms and smaller.

X - Average.
                            70

-------
                                 SECTION  XII

                                  APPENDICES

APPENDIX  A.   PROGRAM- -GRAND OPENING
                                 CROULEV  FOOVS, INC.
            PROGRAM


             9:30 a.m.

            10:00 a.m.
            10:30 a.m.


            11:15 a.m.


            11:45 a.m.

            12:15 p.m.


            1:15 pm
                                          , Hew Vo/tfe
                               GRANV  OPENING CEREMONIES

                                et/  Pnoc.eAi.inQ facitittu
         *  *  * WEPNESPAV, JUNE 21, 1972
 RECEPTION  -  Plant

 WELCOME REMARKS  -
 fnancAA E. Cio(ule.y,
 C>WMle.y food!,, Inc..
                              BRIEF- GENERAL HISTORY ANV FUNCTIONS Of
                              THE LAFARGCl/HLE PLANT -
                              Cnanlu CoAt&on,  Vi.vi44.on Manage*.
                                           Vi.viAi.on
DESCRIPTION OF WHEV PROCESS -
Vn. RobeAt R.  lall,  Consultant

PLANT TOUR - including the. new U/net/
facility

SAMPLING Of POTENTIAL SPECIAL PROVUCTS  -
      whe.y deAi.vati.vu
      to Ale.x.andfii.a Bay, Wew Voik
LUNCHEON - EVGEWOOV RESORT
Aie.xand>ua. Bat/,  Hew Votik

REMARKS -
  Won.  Robert C. McEwen
  Hou&e. 0(! R&pieAe.ntatlv&!>, 3Ut
                                Environmental  Piottction Agency
                                Itlcuklngton, V.C.
                                                                                 I • French
                                                                                   Onion Dip
> Yogurt
                                           71

-------
APPENDIX B.  SAFE OPERATING CONDITIONS
ULTRAFILTRATION SECTION

    Do not exceed 130°F
    Do not freeze
    Do not exceed 55 psig
    Do not permit a vacuum or sudden increase in pressure
    Keep the membranes wet at all times
    Keep pH between 2.5 and 8.8

REVERSE OSMOSIS SECTION

    Do not exceed 90°F
    Do not freeze
    Do not exceed 900 psig
    Do not permit a vacuum or sudden increase in pressure
    Keep the membranes wet at all times

    Keep pH between 2.5 and 8.8
    Limit concentrations of chlorine when sanitizing to
         6-10 ppm of free chlorine
                 USUAL OPERATING CONDITIONS
ULTRAFILTRATION SECTION

    Temperature:  120-125°F
    Pressure:     50 psig at the highest point
                  15 psig at the lowest point

REVERSE OSMOSIS SECTION

    Temperature:  80-85°F
    Pressure:     800 psig at the inlet to the modules
                           72

-------
APPENDIX C.  CLEANING AND SANITIZING SOLUTIONS

ULTRAFILTRATION

Cleaning

Ultra-Clean*:  Part "Au is powder which is dissolved in warm
water (110-112°F) as indicated on the label.  The pH is
adjusted to 7-7-8.5 with caustic or phosphoric acid as
required.  Part "B!l is added after the "A" material is
dissolved.  Rate of additions are "A" at 42 lbs/1,200 gals.
and "B1- at 6 lbs/1,200 gals.,
or
Alcgzy_ne**:  Use at one ounce per 2 gals, of water at
110-120°F  (37-1/2 lbs/1,200 gals.).  pH is adjusted between
8.0-8.7 with phosphoric acid or sodium hydroxide.  The
Alcozyne powder is added to preheated water just prior to
use .

Cleaning solutions are used in the concentrations resulting
from the mixtures stated above.

Sanitizing

The UF solution consists of one ounce of Antibac B+ per 12
gals of water at room temperature, which provides 100 ppm
free chlorine (6-1/4 Ibs in 1,200 gals).  The solution is
pumped through the system just prior to whey operation and
is not used for storage purposes or a
RoccaJL"1"1" solution is circulated through the system at 200
ppm concentration.  This mixture is made up at the rate of
1 oz/4 gals or 1,000 gals required 250 oz of materials.  In
addition;  phosphoric acid is added to adjust pH to 4.5-5.0
at the rate of 1.5 mis/gal of solution.  The 1,000 gals of
sanitizing solution requires 1,500 mis.  Acid addition
increased  the biocidal efficiency of this sanitizer.

REVERSE OSMOSIS

Cleaning

Emb_iozy_me_RQS-j/-   at the rate of 10 cc/gal of water (6 lit
or 1-1/2 gals in 600 gals of water).  Solutions must be used
at once and not stored.  -The use temperature of the cleaner
is at 85-90°F.
                             73

-------
Sanitizing

One ounce of Antibac B per 120-200 gals of water at room
temperature and a pH of 5.5-6.5.  This mixture provides a
solution that will contain between 6-10 ppm of free chlorine
and is pumped through the system just prior to its use for
processing of permeate.
* Abcor,  Inc., 3^1 Vassar Street, Cambridge, Massachusetts
02139
**Alconox, Inc., 215 Park Avenue South, New York, NY 10003
+ Wyandotte Chemicals Corp., J.B. Ford Division, Wyandotte,
Michigan  48192
/ Midwest Biochemical, 1500 West North Avenue, Milwaukee,
Wisconsin 53205
                            74

-------
APPENDIX  D.   ANALYSIS  OF COMPOSITE WATER SAMPLES
    Table 1-D.   ANAYLSIS  OF COMPOSITE WATER SAMPLES
                CROWLEY'S LAFARGEVILLE PLANT

                          September 29, 1971
Sample
no.
0


1

2
3
4

5

Alka-
Hard- lin-
Location pH Bess ityb
UF and RO 8.1 340 326
units
(after
filters)
Upstairs 8.3 336 288
culture
room
Tower well g. 1 356 286
Boiler 8.2 364 291
room well
UF and RO 8.3 354 315
units
(before
filters)
Drinking 8.1 342 286
fountain
(bulk cheese
room)
Solids
(%) Cz Zn Fe 504 Si02
422 58 <.01 0.40 42 4.40


363 60 <.01 0.08 62 3.25

364 65 C.01 <.04 66 3.00
358 65 <. 01 <.04 62 3.35
337 55 <.01 .32 41 4.05

443 65 0.22 .04 72 4.70

      Analysis performed by Metcalf and Eddy, Inc., Boston,  Mass.

      mg/1 as CaCO
                               75

-------
               LABORATORY REPORT
CLIENT
METCALF ft EDDY, HNCINEEn?




0
i
n
E.
"-
v;ater
SAf/.I'Lt COLLECTED nY C
DATE RFrn\.'pp -^ — p — / .1 ANAI YilS
ATLCOLLF.Cn tO
Hy iJiJ tJ

C AM

Ub. No.
Sample
Constituant
r-?&76
0
52677
1
5?fi?8
2
!

3
Physical Examination
Color (Phiinum Eurdirdj
Ociot - Co!d
Turbidity

















Chemical Examination Milligrams per liter
pH Value
Free Ccrhon Dioxide as CO2
Alltclinitv - Total as CaC03


Tote! t Isrdntss is C;C03
Chl'.-rids S3 Cl
8.1

326


3 '1C

iron as Fe 1 0.^0
Mar(o3ne'..e r-s Mn
Crth.irihosut'atc 35 K
I Comolex Phosphafd as P
lolai t'Mosntiaie as F



8.3

288


336

0.08 .



|
8.1

286


356

<0 0*i



8.2 1

?Q1
1

^6i| i

<0.0^ 	 J
1
.J
i
1 i
1 1
Sulfatc as S04 1)2.0 i62.0 ! on 0 r>? & \
Hiuitf Kitronen as N J
Nilralc Nitroocn as N
Ammonia Nitrogen BS N'
Organic, i^itro^en RS W
Biocs.emicol Oxyaen Demand
Che-mica: Oxyqen Der.iand
Suspended Solids
Volatile
Total SoMds
Volal.le
Calcium an Ca
Zinc an 7n
Silica as S102






i





1]22

'"-R
-'0.01










36?

60
cO. 01
^.25









7 6 'l









•J^ft

-so o^ to.oi ,
3.00

3.35 !

1 ! 1
i


I
I
Bact'.ria! Ej:cmin:tion
Bactrria per ml on As:: "ir. hr.-3D°C
Bacc.-ti.T p™ ml on Anrr <"li :v.-?0°C
Coli-ccico^ocs orojp
1 ml
C. 1 ml
Confirmed 7i-l (l:G)
Wc-S 1'iobiSlc ,'Nliirnbcr ;;w lO.'i m!
{.o'lfnni.v'IOQml. MF






















i r '" \ j s
I

i

i

j

•Nl/I.UfcaT.'ETSrOI.iTIVS /TOTAL IK'^CLR TLSTS ^ . , , /t-'i'"
« IIT:I irp pv {^V.- lv».v 'Y^J A- r Jr'/ »ivr 10')^ ~"? '
                      76

-------
APPENDIX D
                         LABORATORY REPORT
   CLItNT.
   ANALYSIS nr  water
   SAMPLE COLLECTED BY.




   DATE RECEIVED	
                         DATECOLLtCTCD.
10-5-71
                                      ___ ANALYSIS Bv
                                     J C
m
c
u
5
z
m
O
D
•B
u.
J
<
r - Cold
Turbidity

















Chemical Examination Milligrams per liter
pH Value
Free Carbon DioxWe as CO?
Alkalinity - Total es CcCO3


Total Hardness as OCO3
Chloride n Cl
Iron as Fe
f/.anqanese as Mn
Orthophosphate as P
Complex Phoiphate as P
Toial rhoipnate as r
Suli'ateasSO4
Nitrite Nitrogen as N
Nitrate Nitrrwjert as N
Ammonia Nitrogen as N
Cfnanic Kilrccen as N
Biochernical Oxygen Demand
Chemical Oxypen Demand
Suspended Solids
Volatile
Total Solids
Volatile
Calcium as Ca
Zinc as Zn
Silica as S102




8.3

315


354

0.32




41.0








337

rjr;
:0. m
4. OS




8.1

286


342

0.04




72.0








443

6^
0 ??
4. 70
































































Bacterial Examination
Bacteria per ml or. Aqar ?!
Conli.mtd Test (BG)
Most Prot'sbii- humbcr per 1UO m!
Cotiiorms/ICX) ml. Ml-
















1






O f'J




'


   •KUMBfiB TESTS POSITIVE /TOTAL NuWPtn T£STS
                                    crntiFi
                    re DY _Q^i^L^.t J-'&Vl JI>_lfilJj
                                     77

-------
APPENDIX E.  TUBE FAILURES
Table 1-E.
            TUBE FAILURES IN UF CABINETS DURING A 1972-74
            PERIOD PRESENTED IN CHRONOLOGICAL ORDER
                     Location
                    by cabinet
                                                Failure
Date
19.12
6/1
6/6
6/15
6/26
6/28
6/29
6/30
7/28
8/1
8/8
8/10
8/18
8/22
9/1
9/11
9/18
9/20
9/23
10/2
10/6
10/7
10/10
10/13
10/17
10/20
10/20
10/21
number

1
2
1
1
1
1
1-3
5
1
1
4
5
3
3
1
4
5
1
2
2
1
4
1
5
6
4
4
ty_p_e -

leaker
rupture (2)b
broken end
broken end
rupture
broken end ( 2)
rupture (2)
rupture
rupture
broken end
weld break
weld break
broken module
rupture
rupture
rupture
rupture
rupture
rupture
rupture
rupture
rupture
rupture
rupture
broken end
weld break
rupture
aDescription of various failure types listed at end of data
^Number of failures,  if more than one
                            78

-------
Table 1-E (Cunt.)
TUBE FAILURES IN UF CABINETS DURING A
 1972-74 PERIOD PRESENTED IN
 CHRONOLOGICAL ORDER
     Date
     10/27
     10/31
     11/4
     11/8
     11/3
     11/11
     11/13
     11/15
     11/20
     1 1/22
     11/27
     1 1/29
     11/29
     12/2
     12/2
     12/4
     12/13
     12/13
     12/13
     12/16
     12/18

     1273
      1/8
      1/13
      1/17
      1/22
      1/22
      1/27
      1/29
      1/30
      2/2
      2/2
      2/2
      2/2
      2/2
      2/3
      2/3
      2/5
      2/6
Location
by cabinet
number
2
4
6
3
2-3
1
1
6
2
5
3
2
6
3
2
2
2
3
4
5
3
6
2
4
1
4
6
1
2
1
5
4
6
4
1
3
1
1
Failure
type
rupture
rupture
rupture (2)
rupture
rupture
rupture (2)
rupture
rupture
rupture
rupture
rupture (2)
rupture
rupture
rupture
rupture
rupture
rupture
rupture
rupture (2)
rupture
rupture
rupture
rupture
rupture
rupture
rupture (2)
rupture
rupture
rupture (2)
rupture (3)
rupture
rupture
rupture (2)
split (3)
rupture (2)
rupture
rupture
rupture
                             79

-------
Table 1-E (Cont.
TUBE FAILURES IN UF CABINETS DURING A
1972-74 PERIOD PRESENTED IN
CHRONOLOGICAL ORDER
Location
by cabinet
Date number
•2/6
2/6
2/7
2/16

Feb. - Plant shut down
3/16
3/19
3/26
3/27
3/31
4/3
4/10
4/11
4/13
4/17
4/20
4/20
4/23
4/24
5/2
5/4
5/4
5/4
5/5
5/5
5/7
5/8
5/9
5/12
5/15
3
2
4
1-2

for retubing
2
4
6
3
4
3
6
4
2
6
5 ,
3
1
4
3
1
1
2
1
1
6
4
6
6
3
Failure
type
rupture (2)
rupture
bad end
numerous off
line leakers

rupture
weld break
rupture
rupture
weld break
weld break
rupture
rupture (2)
rupture
rupture
rupture
weld break
weld break
rupture
weld break
rupture
weld break
weld break
weld break
rupture
rupture
weld break
rupture
rupture
rupture
Replaced 36 UF modules from rows 1  & 2
with 36 new type fibergalss-backing.
      5/18
      5/21
      5/21
      1
      1
      5
                    .n Cabinet #1  unit
rupture
leaker
leaker
                             80

-------
Table 1-E (Cont.)
TUBE FAILURES IN UF CABINETS DURING A
1972-74 PERIOD PRESENTED IN
CHRONOLOGICAL ORDER
     Date_
      5/22
      5/23
      5/26
      5/28
      5/29
      5/30
      6/1
      6/2
      6/5
      6/12
      6/13
      6/18
      6/18
      6/19
      6/19
      6/20
      6/23
      6/29
      6/29
      6/30
      7/3
      7/3
      7/6
      7/7
      7/9
      7/10
      7/10
      7/16
      7/17
      7/18
      7/21
      7/21
      7/23
      7/23
      7/24
      7/24
      7/24
      7/24
      7/25
   Location
  by cabinet
  _number	
      1
      6
      2
      4
      4
      3
      3
      2
      5
      6
      4
      3
      4
      4
      2
      2
      1
      1
      4-
      4
      6
      2
      5
      3
      5
      1
      3
      6
      4
      6
      4
      3
      4
      5
      4
      3
      3
      6
      1
    Failure
	ty;p_e	
 rupture
 rupture
 weld  break
 rupture
 rupture
 weld  break
 rupture
 rupture
 rupture
 rupture
 weld  break
 weld  break
 rupture
 rupture
 rupture
 rupture
 weld  break
 weld  break
 weld  break
 weld  break
 rupture
 rupture
 weld  break
 weld  break  (2)
 rupture
 rupture
 weld  break
 weld  break
 weld  break
 weld  break
 weld  break
 weld  break
 weld  break
 rupture
 rupture
 weld  break  (4)
 rupture
 weld  break
 weld  break
                            81

-------
Table 1-A (Cont.)  TUBE FAILURES  IN UF  CABINETS  DURING  A
                   1972-74 PERIOD  PRESENTED  IN
                   CHRONOLOGICAL  ORDER
Date
7/25
7/27
8/1
8/3
8/4
8/4
8/7
8/7
8/7
8/7
8/7
8/11
8/11
8/13
8/13
8/13
8/14
8/15
8/17
8/20
8/20
8/20
8/21
8/22
8/24
8/24
8/25
8/27
8/27
8/28
8/28
8/29
8/30
8/30
8/31
8/31
8/31
9/3
9/3
Locat ion
by cabinet
number
5
2
5
4
6
3
5
3
6
6
2
5
2
4
5
2
4
4
5
2
2
5
6
S
5
4
4
5
2
5
1
6
4
5
4
4
3
6
1
Fai lure
type
rupture
weld break ( 2 )
weld break
rupture
weld break (2)
weld break
rupture § weld break
weld break (2)
rupture
weld break (2)
rupture
rupture
rupture
weld break
weld break
rupture
weld break
weld break
weld break
rupture
weld break
rupture
weld break
weld break
rupture
weld break
weld break
rupture
weld break
rupture
rupture'
weld break
rupture (2)
rupture
rupture
weld break
weld break
weld break (2)
rupture
                              82

-------
Table 1-E (Cunt.)  TUBE FAILURES' IN UF CABINETS DURING A
                   1972-74 PERIOD PRESENTED IN
                   CHRONOLOGICAL ORDER
Date
9/4
9/5
9/5
9/6
9/7
9/7-
9/10
9/10

9/10
9/10
9/10
9/10
9/11
9/11

9/1 1
9/12

9/12
9/12

9/13
9/13

9/14
9/17
9/18

9/19
9/20

9/24

9/25
9/25
10/1
10/1
10/1
10/1
10/1
10/2
Location
by cabinet
number
1
2
3
6
4
6
6
4

3
4
4
1
4
4

4
1

3
1

2
1

5
4
1

2
3

1

6
5
4
5
5
4
3
5
Failure
ty;p_e
rupture
leaker
rupture
weld break
rupture
weld break
rupture
membrane
separation
weld break
rupture
weld break
rupture
rupture
nipples split
on the end
rupture
membrane
separation
weld break
membrane
separation
rupture
membrane
separation (7)
rupture
rupture
membrane
separation
rupture
membrane
separation (2)
membrane
separation
weld break (2)
rupture
weld break (2)
weld break (2)
rupture
rupture
weld break
weld break
                             83

-------
Table 1-E (Cont.)  TUBE FAILURES IN UF CABINETS DURING A
                  1972-74 PERIOD PRESENTED IN
                  CHRONOLOGICAL ORDER
Date
10/5
10/6
10/8
10/8
10/9
10/10
10/10
10/10

10/12
10/12

10/12
10/12
10/12
10/12
New Fiberglass
New Fiberglass
10/13
10/15
10/15
10/15
10/16
10/20
10/20
10/20
10/22
10/23
10/26
10/27
10/29
10/29
10/30
11/2
1 1/2
11/3
11/3
11/3
11/3
11/3

Location
by cabinet
number
6
4
3
2
4
3
5
2

6
2

3
5
4
5
modules Cabinet
modules Cabinet
6
6
6
2
4
4
5
2
3
3
5
4
3
3
4
6
3
4
6
6
3
4

Failure
tj[p_e
rupture
weld break
weld break
membrane separation
rupture
membrane separation
membrane separation
membrane separation
(8)
weld break
membrane separation
(2)
membrane separation
membrane separation
rupture
rupture
#1 UF units
#2 UF units
weld break
rupture
weld break
leaker
rupture
rupture
weld break
weld break
weld break
rupture
rupture
rupture (2)
rupture
membrane separation
weld break
membrane separation
rupture
weld break
rupture
membrane separation
membrane separation
membrane separation
(18)
     11/7
weld break (2)
                            84

-------
Table 1-E (Cont.)
TUBE FAILURES IN UF CABINETS DURING A
 1972-74 PERIOD PRESENTED IN
 CHRONOLOGICAL ORDER
Date
11/7
11/7
11/8
11/8
11/9
11/9
11/9
11/12
11/12
11/12
Replaced
11/12
11/13
Replaced
11/14
11/14
11/16
11/16
11/16
11/16
11/20
1 1/20
11/26
11/26
12/12
12/13
Replaced
12/21
We took
12/22
~1/T8
Put
2/13
3/12
3/13
Location
by cabinet Failure
number type
3
4
4
4
6
5
3
3
6
4
with new type
6
4
F row with new
4
1
5
5
3
3
3

3
3
4
4
membrane separation
weld break
weld break (11)
membrane separation
weld break
membrane separation
membrane separation
weld break
weld break
membrane separation
fiberglass modules
membrane separation
membrane separation
type fiberglass modules
membrane separation
weld break
rupture
membrane separation
membrane separation
weld break
weld break
membrane separation
rupture (2)
membrane separation
weld break

71 modules with fiberglass in rows H, I,
3
out all of the
1
6
new fiberglass
1
2
1

modules in unit #3 = 216
fiberglass rupture


membrane separation
membrane separation
membrane separation
(3)


(7)


(16)


(18)

(14)
(8)

(5)






(18)

(4)


J, K








                             85

-------
Table 2-E.   TUBE FAILURES IN RO CABINETS DURING A 1972-73
             PERIOD PRESENTED IN CHRONOLOGICAL ORDER
Date
19.72
6/7
8/1
6/2
8/7
8/9
8/10
9/6
9/.13
12/2
12/1 1
12/27
12/27
1273
1/3
1/3
1/10
1/15
1/17
1/20
1/20
4/13
7/28
8/25
Location
by cabinet
number

4
2
4
3
2
4
4
3
4
1
2
6

2
4
4
1
2
6
3
2
1
6
Failure
ty&e3-

broken module
broken tube
rupture
broken tube
broken module
broken membrane
borken module
pin hold in module
rupture
rupture
rupture
tube came unglued

loose screw
loose screw
rupture
rupture
rupture
rupture
broken screw
rupture
rupture
rupture
     Description of various failure types listed at end of
     data.
                             86

-------
Table 3-E.  Failures in Ultrafiltration Tubes

      Types                       Descript ion

Broken end	Membrane or its porous support separates
                        from the tube's plastic threaded tip
                        end .
Broken module,
.A  cabinet  containing  216
 off  the  flow  stream  line
 problem.
tubes is taken
for operating
Leaker
Membrane separation
Rupture
 Split	

 Weld  break
 Permeate  clarify  becomes  cloudy  as  the
 membrane  fails  to retain  pore  size  and
 allows  protein  seepage  to occur.

 Membrane  loosens  from backing  like
 bubbles in  the  tube  interiors  which
 tend  to rupture over time.   Occasionally
 a  membrane  section separates enough to
 appear  as a thin  tube within a tube.

,A  weakened  site on the  membrane  support
 structure breaks  like a tire blowout.

.Support tube splits  along its  length.

,The ten-feet UF tube was  made  by join-
 ing two five-feet porous  supports
 together  by heat  welding  and then cast-
 ing a membrane  on the ten-feet length.
 Periodically these tubes  come  apart at
 the weld  site spoiling  both  tube and
 membrane.
                              87

-------
Table 3-E (Cont.)  Failures in Reverse Osmosis Tubes


    Type

Broken module	Herein refers to a faulty unit which
                        contains 20  5-ft long tubes into which
                        the RO membrane is inserted.  The unit
                        fails, not a single tube or membrane,
                        most usually when the screws in its
                        metal cap break.

Broken tube	Tube splits, usually along its entire
                        length which allows RO liquid to gush
                        from the tubal fracture.

Loose screw	Product seeps slowly from either top
                        of module bottom.  The defect is
                        noticed by a permeate color change from
                        water clear to pale yellow.

Rupture	Membrane support ruptures at a weakened
                        spot and allows RO liquid to spurt
                        from a ragged hole usually about an
                        eighth inch in diameter.  The failure
                        is violent as the system iss under about
                        600 psi.


APPENDIX F.  LOG AND NOTE FOR THE OPERATOR

The operator will probably best be satisfied if he keeps a
journal in which he religiously records everything he
notices including the time and date on which he notices it.
In this case, a column for "Comments" on the log sheet can
be used to cross reference events to a formal journal.

Crowley's elected to set up separate log forms for selected
activities.   These are RO, UF, and analytical services.
Examples of important reports to monitor the plant in actual
operation are attached.
                               88

-------
The following information is logged daily:

     Date
     Time
     Comments
     System inlet pressure
     System outlet temperature
     Feed Temperature
     Feed flow rate UF
     Feed flow rate RO
     Outlet (concentrate) flow
     Permeate conductivity for RO
     Permeate color clarity for UF
     Concentrate total solids
     Whey volume as silo level

The compositional and bacterial information logged are

     Raw Whey
     UF Permeate
     Concentrate Whey
     RO Concentrate
     RO Permeate
     RO Permeate (Compositional Sample)
     Protein Concentrate
                          89

-------
UNIT PROCESSING FACILITIES
                                                                    LAFARGEVILLE, NEW YORK
UF LOG
Time of Day
Elapsed Time
P (High) PSI
P (Low) PSI
Temp, (in loop) °F
Silo Level (in gal.)
Balance Tank Process
Sample No.
Silo Temp.
pH Whey
2300


00
c 	
•H
N
•H
^J
•H
C
ffi
t/1



2400
0
40
8.5
113
4500


127
4.5
0100
1
43
11.5
115
3800


128
-
0200
2
44
11.5
115



_
-
0300
3
44
11.5
116
3000


130
4.5
0400
4
45.5
12.5
114
~T4
2000
4

130
-
0500
5
46
13
112
#4
2000
9

134
4.5
0600
6
46
13
114
i?4
2000
9

134
-
0700
7
45
12.5
114
2400
9


4.5
0800

f.
3
,_i
o
p C
Is it
O 0)
•0 rH
4J
3 T)
c/3 nj

0900









1000









RO LOG
P ( IN)
P (OUT)
Temperature
Flow (IN)
Flow (OUT)
Total Cond.
Conductivity #1
Conductivity #2
Conductivity #3
Conductivity #4
Conductivity #5
Conductivity #6
Sample Material




Ml
«H
N
4-1
•H
to
C/3





770
440
86
27.5
7
Spray







790
480
88
26.5
6.6
i-ng







790
540
L- 88J
24.0
6








790
540
88
23.0
5.8
8.6







790
570
85
22.0
5.4
9.4







790
_560_
86
22.0
5.6








790
u 56°
92
22.0
5.6
9.6







790 i 800
800
580j 600 [ 600
90
21.5
5.1
10.0







89
21.5
5.1
_







89
21.5
5.1








800
600
90
21.5
5.1








NOTES:
                                            90

-------
Crowley Foods, Inc.
Whey Processing Facilities
Request for Analytical Services
         LaFargeville, New York
Date sampled_
Date processed_
Sample
No.
1
2
3
4
5A
5B
6
Description
Raw Whey
UF Permeate
Concentrate
Whey
RO Concentrate
RO Permeate
Call Sample
RO Permeate
Protein
Concentrate
Time of
Sample
2230
0200
0230
0230
0230
0230
0430
0530
Tests Required
Solids
5.95
5.43
6.64
16.14
.62

8.16
Acid
.50
.45
.55
1.18
.10

.65
pH
4.5
4.45
4.5
4.55
4.2

4.45
Bacteria
Coli

70
-20
370
20

-10
Yeast
120
40
610
730
Cove

620
Mold
NC
NC
310
860
red

NC
SPC
24000
12000
16000
67000
756000

19000

-------
  16




  15




  14




  13




  12





  11




  10
o
o
o


oo
7




6




5




4





3




2
         12345678

                       ELAPSED TIME, HRS




          Figure  1-F.   ULTRAFILTRATION RUN


                    June 15,  1973
                                                 10   11
                           92

-------
o  RUPTURE
*  MEMBRANE
*  END
  5   10
20
15
10

5

-
-
-







* *\ t i
123
1











2











3











•4











c






-
-
-
A ~
-
/
C
u
                                                                   TOP OUT


o
o
o
T



X
O
0 Q
0> °



2
Figure 2-F

O
o
00





O
O





O
D



3 4 5
. UF VIEWING THE NON-MANIFOLD SIDE


O
O
:
«
~
IS
10
5
6 BOTTOM
IN
o RUPTURE
* MEMBRANE
A END
             Figure 3-F.   RO VIEWING  FROM  INLET SIDE
                                   93

-------
                                                    341 VASSAR STREET

                                                   CAMBRIDGE, MASS. OZI39
                                                    TELEPHONE (617) 401-6640

                                                      TWX 7IO32OO446

                                                        CAULL ADDRrS-..

                                                    ABCOK. CAMBRIDGE. MASS.
APPENDIX G.          CLEANING AMD SANITIZING

            ULTRAFILTRATION&REVERSE OSMOSIS SYSTEMS


  One of the most important determinants of long-term cleanliness
of reverse osmosis and ultrafiltration membranes is the quality
of the water used to clean and flush them.   It must be free of
colloidal matter, especially colloidal iron.  Fine rust particles
will deposit on and adhere to membranes so  tenaciously that their
removal is extremely difficult.   Other colloidal material normally
found in water will deposit readily on membranes, especially on
high flux ultrafiltration membranes.

  For these reasons, Abcor strongly recommends that all water
used to flush or clean membranes be filtered to remove colloidal
material, and be chemically treated to remove iron if that is
necessary.


I.  ULTRAFILTRATION

    General Instructions.  The unit should  be maintained at the
    running temperature during the flushing and cleaning cycle.
    Do not allow the unit to cool down at any time between the
    run and the end of the cleaning cycle.

    A.   Product Removal.   The product must  be pushed out with
        water that is at  the run temperature.   Any instructions
        to the contrary should be amended to agree with this
        instruction.

    B.   Cleaning.

        1.  Flushing.   Before cleaning the  system with cleaning
            solution,  it  should be thoroughly flushed with fresh
            water which should pass once through the system and
            out to drain.  Use a modest flow rate (5-15 gpm per
            parallel pass) and lowest possible outlet pressure.
            If the system has a circulating loop and separate
            feed pump,  then only the feed pump is turned on while

                                                    PIB 210
                                                    Rev. 8/23/73
                                94

-------
flushing.  The valvincj should be arranged so  that
the water par: see oncti through tho system and  does not
circulate iiround the loop.

Flush a volume of water which is between 1.5  and 3
times the internal volume of the system, an.d  is at
the temperature of tho run.  At the conclusion of the
flush, the cleaning solution concentrate is pumped
directly into the flushwater left in the system.
        ^ S_oluJ-J .on •  To clean ultraf iltration systems,
the cleaning solution should contain a proteolvtic
enzyme.  These enzyrr.es lose their activity quite
rapidly once they have been dissolved in warm water.
Cleaning solutions should be prepared just before use
and should not be held.  Two different solutions are
.available :

ABCOR ULTKACLKAN .   U] traclean is shipped in two parts,
Part A which is  the detergent and Part B which is the
enzyme.  Dissolve Part A at the rate indicated on the
package, or at a  rate of 0.42 lbs/100 Ibs of final
cleaning solution in the system.  The volume of the
water remaining  in the system from the rinsing step
must be included  in the calculation of the total water.
Use water at least as warm as the run to dissolve
Part A.  Do not  exceed 55°C  (130°F).  Make sure that
Part A is completely dissolved.

Add Part 13 at the last possible moment before this
mixture is pumped into the system.  It is hot absolutely
necessary that Part Js be completely dissolved before
added to the system  (but it ir, necessary that Part A
be totally dissolved) .  Part B can be added after the
detergent solution is pumped into the system, but the
delay should not exceed five minutes.

After thorough mixing with the water already in the
system, monitor  the system pH.  It should be in the
range of  7.7 to  8.7.  If the pH is outside this range,
add phosphoric acid or sodium hydroxide as indicated
to the concentrate before starting the next cleaning
cycle.  You should quickly discover how much of acid or
alkali is needed to maintain the proper cleaning pH.

ALC02YHE.*  Dissolve Alcozyme at a rate of  .37 lbs/100  Ibs
'of water  in the  final cleaning solution making a con-
centrate  as above.  The solution must be used immediately,
as the enzyme cannot be separated from the detergent
in this preparation.

Monitor the pH as indicated above with Ultraclean, and
nntc that it will generally be necessary tc add
phosphoric acid  to maintain a pit below 8.7.
                        95

-------
If your system has a circulating loop and separate
feed pump connect the cleaning solution tank to the
suction side of the feed pump just as you would
connect the whey tank to the suction side of the
feed pump when running whey.  Connect the return
line to the cleaning tank instead of to the whey tank.

Cleaning Procedure.  Start the system very much the
way you would start a whey run.  Start the feed pumps,
then start the circulating pumps and establish full
running circulation in the system.   Open the return
valve all the way, circulating with the lowest possible
pressure in the system.  It is important that the
velocity through the tubes during cleaning be at least
as high as the velocity through the tubes during
running, and that the pressure at the outlet of the
unit be lower.

Monitor the flux from the unit.  This is of critical
importance.  The flux will rapidly climb, then plateau,
then drop.  The plateau may be very brief, and it is
important to locate it.  As soon as the plateau is
reached, dump the cleaning solution and immediately
rinse the unit with water.

Rinsing.  As soon as the system is  cleaned  it should
be thoroughly rinsed by pumping a volume of water
equal to 3 to S times the internal  volume of the
system, once through the cysLera, and out to drain.
This water may be at any temperature below 50°C.
During this rinsing step, check the water flux, cor-
recting for pressure and temperature.  Compare it with
a recorded running average of water fluxes.  If the
water flux in the unit is not up to the running average,
immediately reclean the system with warm water, and
10% of the cleaning agent level indicated for normal
cleaning.  Water fluxes may be run with water or saniti-
zing solution, but not with any other  agent present.

Modifications.  This cleaning procedure is designed to
give very fast cleanups.  Frequently,  the actual detergent
washing cycle is 10 minutes or less.   It is  very important
that tliis step be cut off r.t the flux  plateau;  overcleaning
results in lower fluxes than optimal cleaning.

Once good cleaning has become routine, one may attempt
to reduce these detergent quantities somewhat.   Levels
have been reduced to 0.7 gm/square  foot of membrane
area in some well run plants.  That level should be
approached with caution.
                      96

-------
C.  Sanitizing
        Sanitizing Solution.  ANTIBAC B***.  Dissolve
        Antibac B at the rate of  0.03 lbs/100  Ibs of water
        at betv;een 60° and 90°F — 15°-35°C, and adjust the
        pH as may be required to  be between 5.5 and 6.5.
        This solution has 50 ppm  free chlorine.  Check the
        free chlorine content as  the sanitizing is being
        completed.  If this is not above about 20 ppm,
        then you have been sanitizing an unclean system.

        Sanitizing Procedure.  Fill the system with sanitizing
        solution, being sure to displace all the water that
        was inside the system.  Circulate gently using feed
        or circulation pumps for  between 10 and 15 minutes.
        Be careful of the timing.  Do not exceed 15 minutes.
        Do not sanitize for Ions jthan 10 minutes.  Be sure
        that every pump is started and that every valve is
        opened and closed for a thorough sanitizing of the
        system.  Adjust the pressure so that about half the
        volume of sanitizing solution being pumped into the
        apparatus' leaves it as permeate, if possible.  This
        will sanitize the porous  backings of the membranes.
        The minimum amount of permeated sanitizing solution
        is 1 gallon for every 10  ft2 of membrane area.  Routinely
        check the free chlorine concentration  in the permeate.
        It should be close to the membrane side concentration
        at the end of the sanitizing run.  If  it is significantly
        lower, the porous backings are probably dirty, and this
        js the first warning you  will sec.  Discard permeated
        sanitizing solution.

        Filling the System.  Do not leave sanitizing solution
        in contact with the membranes for longer than 15 minutes.
        As soon as sanitizing is  complete, displace the
        sanitizing solution with  either water  or whey .  It is
        customary to sanitize just before starting up and  to
        displace the sanitizing solution with  whey.  This  is
        satisfactory if the shut-down period between runs  is
        not  longer than 6 hours.  If the system is to be idle
        longer than 6 hours, you  should sanitize when you  shut
        down and again when you start up.

        Cleaning Membrane Exterior .  The outside of the membranes
        may  be sprayed with water and cleaning solution the
        same as used inside the membranes.  The outside only_  of
        the membranes may be sanitized with sanitizing  solution
        up to 200 ppm chlorine.   However, be very  careful  that
        this sanitizing solution  does not get  inside  the  syctem.
        Do not leave the outsides soaking in  this  strong
        solution.
                                97

-------
     D.   Storage.

         If the time between operating usage is hours or up to
         3 days,  as in normal plant cycles,  execute the cleaning
         procedure, sanitize and then fill the membrane system
         with filtered water at room temperature.   Sanitize again
         before start-up.   If the standby time is  as much as a
         week, clean and sanitize,  and allow a trickle of water
         to continuously flow through the system.   For extended
         periods,  the system should be continuously purged with
         filtered water and cleaned and sanitised  weekly.  A
         residue  of "Zcphiran"'1"1" sufficient  to inhibit
         microbiological activity is recommended.
II.   REVERSE OSMOSIS

     A.   Interior of the Membranes

         All cleaning of reverse  osmosis  systems  should' be done
         at the lowest possible pressure.   This usually means
         physically disconnecting bucv.  pressure regulators.   Also,
         in many reverse osmosis  systems  the  residence time of
         a liquid being pumped through  the  system is longer than
         the normally permitted contact time  between cleaning
         solution or sanitizing solution  and  the  membranes.   This
         means that we will  be changing the solution at the feed
         end of the system before this  solution is seen to be
         leaving the system  at the  exit.  Cleaning is done by
         time, not by examining the solution  leaving the system.

         1.  Flushing^.  Before using cleaning solution, flush the
             system by pumping through  clean, filtered water.
             Pump a volume of water which is  equal to 3 to 5
             times the internal volume  of the system.

         2-  Cleaning Solution.   EMBIOZYME  ROS-1**.   This is an
             enzyme cleaner.   The enzyme  will deteriorate quite
             rapidly when the cleaner is  diluted  with water, so
             prepare the cleaning solution  just before it is to
             be used.  Dissolve Embiozyme ROS-1 at the rate of
             10  ml/gallon  water at  85°F (30°C) and use immediately.
                                98

-------
         _____.            cleaning solution through
        the system for 30-minutes.   Watch the liquid, leaving
        the system.  When this liquid is cleaning solution
        and not water, then it may be returned to the
        cleaning solution tank and be re-circulate^.

        Rinsing.  Aftcf cleaning, pump fresh', filtered
        water once through the system and out ,to drain.
        Pump a-volume of water equal to between 3 and 5 times
        the internal volume of the system.  It is most important
        to rinse out all the cleaning solution, because this
        will react with sanitizing solution and lower the
        activity of the sanitizing solution.

        Sanitizing Solution.  ANT1BAC B***.  Dissolve one
        ounce of Antibac B in 150 gallons of water at room
        temperature.  Check the pll which should be between
        5.5 and 6.5.  If the pll is high, add phosphoric
        acid.  If the pll is low, add caustic soda.

        Sani ti zing Procedure.  Pump sanitizing solution
        once through the system and out to drain.  Pump for
        exactly 10 minutes.  Dp_ii°_t exceed J 0 minutes.  Dp
        not use stronger solution than specified. Sanitizing
        solution should  have between 6 and 10 ppm free
        chlorine, not more.  As soon as 10 minutes have
        passed, switch the suction of the pump to fresh/
        filtered water or to whey, and pump until the solu-
        tion leaving the system has less than 1 ppif. free
        chlorine.  Never leave sanitizing solution sitting
        inside the system.
B.  Membrane Exteriors
    When rinsing, cleaning or sanitizing the outside of
    RO membranes, there must be a pressure inside the
    membranes of not less than 10 psi.  This iu necessary
    to prevent the chemicals used on the outside of the
    membrar.es from getting at the inner surface of the
    membranes.  If, however, the inside pressure is too
    high, then the permeation rate will be such that
    the chemicals will not make adequate contact with
    the outside of the membranes.  To insure that there
    is a correct pressure inside the system, it is only
    necessary to insure that liquid is, in fact, flowing
    into and"out of the system.  In small systems, direct
    connection to city water will be adequate.  In larger
    systems, the booster pump is usually kept running.
                            99

-------
        No solution which has contacted the outside of the
        system should over be permitted contact to the inside
        of the system.   There are two reasons for this.  It
        is most important that micro-organisms that may grow
        on the outside not be introduced inside the system.
        It is also desirable to use harsher chemicals to sanitize
        the outside than are ever permitted inside the membranes.

    1.   Flush i_ng.   Before cleaning, flush the outside of the
        membranes  with water for a few minutes.   Inspect the
        membranes.  If growth is seen on the membranes, remove
        it with a  jet from a cleaning hose.

    2-   Cleaning.   Clean the outside of the membranes by spraying
        15 to 20 minutes with the same solution  that has been
        used to clean the inside of the membranes.   The solution
        may have already been used to clean the  inside of the
        membrane.   However,  once this solution is used to clean
        the outside of the membranes, it must be discarded.  Do
        not permit it to go inside the membranes.

    3   Sanitize.   Sanitize with any available sanitizing solution
        (for example, hypochlorito or Antibac B), using a solution
        of about 100 ppm chlorine.   Spray the outside of the
        membranes  for about 30 minutes.   If growth has been
        observed on the outside of the tubes,  use 200 ppm chlorine
        solution several times each week.   Tank  cleaner may also
        be used on the  outsides if the precautions  listed above
        are scrupulously followed.

C.  Storage

    Same application as ultrafiltration.
                               100

-------
  +Ultraclean,  Abcor,  Inc.   341  Vassar St.,  Cambridge,  Mass.

        ULTRACLEAN  is  a proprietary  sanitary  cleaner for use with
        Abcor membranes used for  dairy  and other sanitary applications.

        Usage

        1.   Determine  water  hardness and iron content.   Note:   If
            iron content is  higher than 0.2 ppm, the water must be
            pretreated to lower the  iron content.

        2.   Determine  amount of Part A  of ULTRACLEAN to  use per
            100  gallons from the  table  below:

              Calcixim Carbonate            Quantity per 100
              Water Hardnesn              Gallons  Vtfater
              Parts per Million            Part A	Part B

              up to 350                   38 Ibs.      3 Ibs.

            For  hardness ab_oy_c 350,  increase  Part A  in proportion
            to hardness.  For hardness  below  350,  use-as above.
            For  very large quantities,  contact ABCOR for possible
            cost savings by reformulation.

        Add Part B just prior to  use.

        Keep operating temperature between 45° and 50°C.
   *AJ.conox,  Inc., 215 Park Avenue So.,  New York,  New York 10003

  **Midwest Bio-Cheinical,  1500 W.  North  Ave.,  Milwaukee,  Wise.  53205

***Wyandotte Chemicals Corp.,  J.B. Ford  Division,  Wyandotte, Mich.  48192

 +'hZephiran,  Winthrop Laboratory,  90 Park Avenue,  New York, New York 10016
                                 101

-------
APPENDIX  H.     FLOW PATTERN  IN  REVERSE OSMOSIS MODULE
                                  cv
                                 DV
                                        CV
                                                          CV
                                                          CV I
                                                          CV '
                                DV      DV





CV i













1 •*


CV 1








cv I




J—



CV .
CV










:




1




: i
                                DV
ORIGINAL FLOW CONCENTRATION AND
CLEANING
MODIFIED WITH CONTROL VALVES,  REVERSE
FLUSH CLEANING.
MODIFIED WITH CONTROL VALVES,   FORWARD
FLUSH CLEANING.
   CV - CONTROL VALVE,  DV - DRAIN VALVE

-------
APPENDIX I.  METRIC CONVERSION TABLE
Because the Crowley LaFargeville plant used English units
for all purchases, sales, and process measurements, these
are the units used in this report.

The chart below is applicable conversion.
           1 Gallon = 3-784 Liters
           1 Pound  =   .454 Kilograms
           1 Foot   =   .305 Meters
           Degrees F= 9/5C + 32
                           103

-------
                                    TECHNICAL REPORT DATA
                             (Please read Instructions on the reverse before completing)
 1. REPORT NO.

  EPA-600/2-77-118
                                                            3. RECIPIENT'S ACCESSION-NO.
 4. TITLE AND SUBTITLE
   Membrane Processing of Cottage  Cheese Whey
                5, REPORT DATE
                  June 1977  issuing date
                6. PERFORMING ORGANIZATION CODE
 7 AUTHOR(S)

   Robert R. Zall
                                                            8. PERFORMING ORGANIZATION REPORT NO,
9. PERFORMING ORGANIZATION NAME AND ADDRESS
   Department of  Food Science
   N.Y. State College of Agriculture  and Life Science
   Cornell University
   Ithaca, New York  14853
                10. PROGRAM ELEMENT NO.

                  1BB610
                11. CONTRACT/GRANT NO.

                  12060  DXF
 12. SPONSORING AGENCY NAME AND ADDRESS
   Industrial Environmental Research Lab--Cin., OH
   Office of Research and Development
   U.S. Environmental Protection Agency
   Cincinnati,  Ohio 45268
                13. TYPE OF REPORT AND PERIOD COVERED
                  Final	,	
                14. SPONSORING AGENCY CODE
                  EPA/600/12
 15. SUPPLEMENTARY NOTES
 16. ABSTRACT
        A full-scale whey processing  plant using membranes was constructed  to
   process 300,000 pounds per day of  cottage cheese whey.   The two-step  system
   uses ultrafiltration (UF) and reverse osmosis (RO)  according to a design
   previously  demonstrated in the Phase I portion of this  project and
   reported  in Water Pollution Control  Series 12060 DXF  07/71.   This report
   was submitted in fulfillment of  Grant number 12060  DXF  by Crowley Foods,  Inc.
   under the sponsorship of the U.S.  Environmental Protection Agency.  The
   report covers the period from June 21, 1972 to December 1974, and work was
   complete  as of April 10, 1975.
17.
                                 KEY WORDS AND DOCUMENT ANALYSIS
                   DESCRIPTORS
                                               b.IDENTIFIERS/OPEN ENDED TERMS
  Osmosis
  Chlorination
  Membranes
      Permeate
      Reverse Osmosis
      Membrane Processing
      Ultrafiltration
                                COSATI Field/Group
13/B
 3. DISTRIBUTION STATEMENT
   19. SECURITY CLASS (This Report)
     Unclassified
                                                                          21. NO. OF PAGES
                                                                           114
                                               20. SECURITY CLASS (Thispage)
                                                  Unclassified
                                                                          22. PRICE
EPA Form 2220-1 (9-73)

  •-U.S- GOVERNMENT PRINTING OFFICE: 19"T-"57-056/6461
104

-------