EPA 430/09-87-010
SUMMARY OF THE
1987 CARVER-GREENFIELD
SLUDGE DRYING TECHNOLOGY WORKSHOP:
PROBLEMS AND SOLUTIONS
Held in Los Angeles, CA
on March 10 and 11, 1987
by
:c:-.n Walker, Office of Municipal Pollution Control
and
John Zirschky, ERM-Southeast, Inc.
Contract No. 68-01-7108 with
Environmental Resources Management, Inc.
James Wheeler
Project Officer
Office of Municipal Pollution Control
U.S. Environmental Protection Agency
Washington, D.C. 20460
(4- •
-------
TABLE OF CONTENTS
Page
LIST OF TABLES i ii
LIST OF FIGURES iv
EXECUTIVE SUMMARY V
OVERVIEW 1-1
1.1 Introduction 1-1
1.2 Report Format 1-2
THE CARVER-GREENFIELD SLUDGE DRYING PROCESS 2-1
2.1 Background 2-1
2.2 Multiple Effect System 2-2
2.2.1 Process Description 2-2
2.2.2 Addback 2-11
2.3 Mechanical Vapor Recompression 2-14
2.4 Light-Oil System 2-16
OPERATIONAL EXPERIENCES 3-1
3.1 Operating System 3-1
3.2 City of Los Angeles Hyperion C-C
Treatment System 3-1
3.2.1 Project History 3-1
3.2.2 Cost 3-2
3.2.3 Operational Problems 3~-2
3.3 Burlington Industries 3-5
3.3.1 System Description 3-5
3.3.2 Operational Problems 3-6
3.4 LA County Sanitation District
Pilot System 3-6
3.4.1 System Description 3-6
RECOMMENDATIONS 4-1
4.1 Introduction 4-1
4.2 Responsibilities for Design and
Construction 4-1
4.3 Design 4-4
4.3.1 Philosophy 4-4
4.3.2 Plan and Specifications 4-5
4.3.3 Equipment Selection 4-5
4.3.4 Plant Model 4-7
4.4 Contractor Selection and Construction
Activities 4-8
-------
TABLE OF CONTENTS (Cont'd)
SECTION
4.5 Operations
4.5.1 Personnel Requirements
4.5.2 Operator Training
4.5.3 Start-up and Solids Recycle
4.5.4 Centrate Quality
4.5.5 Excessive Loss of Carrier Oil
4.5.6 Use of Dried Sludge
4.5.7 Process Monitoring and Control
4.6 Innovative/Alternative Technology (i .;
Funding Issues
4.7 Future Meetings
REFERENCES
APPENDIX
A Seminar Attendees
B Reviewers Comments
-------
LIST OF TABLES
.Table Title Page
2-1 Carver-Greenfield Light Oil Plants 2-3
2-2 Summary of United States C-G Light Oil System 2-17
2-3 Summary of United States C-G Light Oil Systems
Operation Information and Energy Requirements 2-18
2-4 Summary of United States C-G Light Oil Systems -
System Redundancy 2-19
3-1 Process Problems and Corrective Measures;
Burlington Industries C-G System 3-7
3-2 Mechanical and Physical Problems
and Corrective Measures; Burlington
Industries C-G System 3-10
i-4-^> FW Scope of Work Carver Greenfield Facility
~" Step II Design 4-2
C^2) FW Scope of Work Carver Greenfield Facility
Step III Services 4-3
-------
LIST OF FIGURES
Figure Title _l£2
2-1 Process Flow Diagram of the C-G System 2-4
2-2 Process Flow Diagram of the MVR/C-C System 2-8
2-3 Schematic Diagram of the Oil Distillation
System for the C-G Process 2-9
2-4 Addback Solids Flow Diagram for the
Multiple Effect C-G Process 2-12
2-5 Process Flow Diagram of the MVR/C-G System 2-15
3-1 Schematic Diagram of the Hyperion Sludge
Process" Facility 3-3
-------
CARVER-GREENFIELD SLUDGE DRYING TECHNOLOGY WORKSHOP:
PROBLEMS & SOLUTIONS
EXECUTIVE SUMMARY
Four municipal wastewater treatment authorities have selected
the Carver-G"°°nf'T1 ^ n ^ ghi" "* "* sludge drying (C-G) svstert
for dewatering their sewage sludge. Ocean County, New
Jersey;theMercerCountyUtilityAuthority (Trenton, New
Jersey); the City of Los Angeles, California; and the Los
Angeles, California, County Sanitation Districts are
currently designing or constructing C-G systems. The City of
Los Angeles will soon begin operation of their 265 dry ton
per day C-G system. This system will be the first municipal
wastewater sludge C-G unit utilizing thp light oil syst-pm to
begin operation. LimifcpH si-art: — up tests have been conducted r
and as with any new application of a t-o^hrmi rvrjy f problems
h_aj£g- been experienced in the start-up of the system. The
problems encountered by the City of Los Angeles have been
aggravated because of the v_pry .ghnri- <-img period ava i 1 a,h1Lf=>
f"r rJP°i7-n—a-»ei—nnnqt-rnnt- i on anH hppausp this sysfprn i_s—tile
first full-scale system to be built.
To minimize start-up problems at the remaining three systems,
the U. S. Environmental Protection Agency's Municipal
Facilities Division (EPA-MFD) sponsored a two-day seminar and
workshop on the C-G process. Representatives from each of
the four wastewater treatment plants, and C-G design firm
(Foster-Wheeler USA Corporation), the patent holder of the
system (Dehydro-Tech), EPA, the state environmental agencies
from CA and NJ, and municipalities currently considering the
C-G process attended the seminar.
This report was written to summarize the considerable amount
of information on the C-G process that was disseminated at
the workshop. The objective of this document is to summarize
the key points of the seminar and other follow-up information
to assist both the attendees and other individuals who may be
considering the C-G process. Suggestions for improving the
design and construction procedures as well as the EPA/state
funding mechanisms are included in the report.
The Carver-Greenfield Process (C-G) is a patented system for
drying solids. It is now being applied in the United States
to the drying of municipal wastewater treatment plant sewage
sludges. The first step in the C-G process is to mix a
carrier (fluidizing) oil with wet sludge (which can have a
wide range of initial moisture contents) . This fluidized
mixture is then fed into a multi-effect evaporation system (a
series of interconnected evaporators and heat exchangers).
The final solids end product is essentially free of oil and
-------
water. The C-G configuration (Figure I) is designed for
efficient utilization and reut il i za tion of heat with the
carrier oil being recovered and recycled repeatedly in the
process. The C-G system is substantially different from any
other wastewater - slndp-p processing systems hprans.'ff i t-
yEj,iiz&s _ recyclable hydrocarbons to fluidize the solids
Hnrina frying. A C-G system, in fact, resembles a
{petrochemical plant more than a wastewater treatment plant
and requires some of the same equipment and _s k i 1 1 s found in
petrochemical plants to properly operate and maintain.
There have been a progression of operating C-G systems for
drying different kinds of solids dating back to 1961. In the
mid 1_9 7 0 s , the_ system wa s modified to a I low n^f nf a 1 igh t
(volatile) fluidizing oil (e.g., No. 2 fuel oil). This
system is the basis for four municipal wastewater treatment
facilities in the United States — the City of Los Angeles
(City of Los Angeles), the Los Angeles County Sanitation
Districts (LACSD), Mercer County (Trenton), KJ and Ocean
County, NJ. Three of these fnnr facilities are employing
four f orced-ci rcul a i- i nn evaporating units in series for
Hrying t-hta q i . n rig. g» r and the fourth — Ocean County, MJ--is using
a mechanical vapor recompression evaporator, followed by two
forced-circulation evaporating units in series for sludge
drying. The first f ul ly oper£tiiQjia_l.^JLgh^
the United Sates is being utilized by Bur linqton^Industr ies
in Clarksvil 1 e_,_ri _V i r gin i a.. • The Burlington system processes to
dryness both wool-scouring wastewater at 0.5 to 1 percent
solids as well as a dewatered solids residue at 16 percent
solids from their wastewater (finished wool cleaning water)
treatment plant.
•The light oil C-G systems must safely pump, evaporate.
para_t^ •qnf^ i-prycle hot oily
_
water and solids through .a.._nnmhpr of i
c o m P Q n_e_n_t-S..^, These task s _have__ posed ^ nnmbejn — o£__dle_JLJJiLP r
co_nstruction, start-up and operational problems -f_or all
qr_oups involved. VaTuable lessons are being learned,
especially from the design, construction and start-up
experiences at the City of Los Angeles. Also, valuable
insjf|hfc naa hpen gained f rom nporat- i nna 1 g xper i gnre-s — o_f _ f.he
jSurlington Industries light oil C-CL sygf-pTn and a pilot pla_nt
testing program under^k ^n in t-ho_iai-g i g_70s_ by .
The U. S. Environmental Protection Agency (US EPA) wanted to
help the municipalities achieve viable C-G systems for cost
effectively drying their sludge. All groups involved (the
designers, manufacturers, and the municipalities) wanted to
find solutions to specific problems such as with construction
management, start-up, grid, and portions of design that were
not working well—the goal being to avoid repeating previous
mistakes and to quickly benefit from design modifications and
vi
-------
The ERM Group.
MOIST
SOLIDS
FLUIDIZING
TANK
CARRIER
OIL
CARRIER OIL
CARRIER OIL
EVAPORATORS
AND
HEAT EXCHANGERS
r
DRIED SOUDS IN OIL
7
SEWAGE OIL
CARRIER OIL
SOUDS
CARRIER OIL
A
WATER
TO
POTW
SEWAGE
OIL
(FUEL)
98% SOLIDS
(FUEL OR
FERTILIZER)
FIGURE I
SCHEMATIC OF THE CARVER-GREENFIELD SYSTEM
Vll
-------
'approaches that have worked well. A workshop offered the
best format for sharing information and improving
communications between all parties. This format also seemed
to be an important way of focusing on difficult problem areas
of the new C-G technology to create an awareness and an
atmosphere for finding and sharing accelerated solutions.
Therefore, a_t-wo-day workshop was held in. Los Angeles in
March 1987.
An excellent spirit of cooperation and willingness to share
among all these groups was established by the workshop.
Communication among the groups has been excellent. This
proceedings of the workshop, which also contains considerable
follow-up information from the participants, has been
produced to summarize the findings and assist other groups
considering utilization of this technology, as well as the
workshop participants. The report is divided into four major
parts as follows: (1) Overview, (2) The Carver-Greenfield
Sludge Drying Process, (3) Operational Experiences, and (4)
Recommendations.
Many problems have been identified and overcome. Other
identified problems are being worked on. These problems
include:
1. OIL-WATER SEPARATOR AND EVAPORATORS: An oil-water
separator and interconnected evaporation system is
used for separation and recovery of the carrier oil
from the evaporated water (not more than about 100
ppm residual .insoluble oil in the water) .
Problems with the oil recovery to date have
included (a) prevention of solids carry-over to the
oil-water separator from the evaporator, and (b)
maintaining reliable operation of the separator.
Solids carry-over at the City of Los Angeles has
prevented formation of a good oil-water interface
in the separator due to the formation of an
emulsig_n. Oil-water separation, however, has also
been inconsistent where solids carry-over to the
separator has not been a problem (e.g., at
Burlington).
If Oil is not properly spparafpd and is carried
over_ into the water, j, t is wagted aacLgoes back_to
the head of the wastewater treatment plant. If
water is not sega_r,ated._.fj:om.__.tbje— oil . the water, is
recycled with the_ ca.rr_i.er oil, to tihe fluidizing
tank and can cause th.e._ formation . of a "gumrnv
phase.'' At a water content in solids and oil in
the range of about 3Q__to 1^ pprr^nt the viscosity
of the solids slurry greatly increases. This high
vii i
-------
viscosity phase is referred to as the gummy phase,
and development of the gummy phase can cause
pumpabilitY problems and plugging iji the
evaporators and heat exchangers. "~~~"-
2. ON-LINE MONITORING OF THE RATIO OF SOLIDS TO WATER
TO CARRIER OIL: Improvement is needed in automatic
measurement and control of the exact amounts of
solids, water and carrier oil bein_g pumped ajd
mixed at critica} Ipnat-inrrs—in t-ho C-G sysfre,IP-
TETiis particularly critical because excessive
carrier oil in the system will reduce the capacity
because of the excess oil that must be volatilized
and recycled. Too llttlp r**rr.i.e*r njj. may likely
result in an oil-solids mixture that is too thick
to pump a solids-water mixture of around 30 percent
(gummy phase) or in which can cause plugging in the
heat exchangers"! "
Manual sampl ing and determination of .._the
solids-to-water-to-oil ratio at half-hour intervals
is being successfully used at the Burlington
Industries C-G facility to aid in control.
3. CENTRIFUGATION: Centrifugation is used after the
evaporative drying process to separate the drv
sol ids from the carrier and sewage oil.
Difficulties experienced with centrifugation have
•-7 included excessive abrasion and insufficient
c_apture of solids fines. Sufficient capture of
fines during centrifugation depends upon a number
of factors including, proper rate of feed and
gravitational force.
Failure to captur-e sufficient solids, sav only 9?
percent as compared wil-h—QQ percent,, mpans that
ther_e will be 3 percent, in.gi-parl oJL. 1. percent fines_
in the sewage and carrier oil going t-n t-hp flash_
still (not yet in use by the City of Los Angeles) .
As the carrier oil is flashed off of the sewage oil
be
eva
pora
t-i
US-
process L_
1-h
p
so.
JdS in_
t-ho_ e:ow
age ...og
1
become concentrated. The resultant solids content
would then 5e about 50 percent instead of 20
percent. This higher solids content can cause
pumpability problems and clogging of the system
piping.
DE-OILER (HYDROEXTRACTOR) : Another problem
encountered by Burlington and now the City of Los
Angeles has been solids cai^ry-over into vapor lines
from the de-oiling| process. Burlington has solved
this problem byincreasing the velocity of the
ix
-------
vapor flow in the pipeline from the de-oiler and by
manually cleaning any accumulated solids out of the
line once each shift (five minutes required per
shift) . The City of Los Angelas ig graying t-i-ia
problem by repipinq to gain an increased velocity
of vapor flow through a shorter, more direct oath.
The City ot Los ft,pgplgs i «=;
installing a cyclonic dust trap with nitrogen
Blanketing as an ext-ra pn=^an<- i rm against solids
bu,.i_ld-up and aui-np* idat-i nn. use of the
distillation system (not yet used during startup by
the City of Los Angeles) should also reduce the
fines in the system.
5. EXCESSIVE LOSS OF CARRIER OIL FROM THE SYSTEM:
Carrier oil losses are thought to occur primarily
(a) into wastewater during oil-water separation,
(b) into the sewage oil from the flash stilling
process, and (c) into the solids product from the
de-oiling process. Excessive loss of the carrier
oil can be expensive"^ Burlington expected a loss
of about 3 gallons per hour but instead has been
experiencing a loss of 9 to 12 gallons per hour.
Depending upon the cost of oil, these losses can be
quite costly. Costs for light oil in August 1987,
were about $1.10 per 'gallon.
PUMPS AND PUMP SEALS: Pnmp g^al failures and
abrasion of pumps has been a continuing problem.
H_j(jh^r gna lit-y pumps with hardened surfaces ar e
jTe.e,d,e,d . Satisfactory service from seals is now
being obtained at Burlington with changes needed
every 3 to 7 months depending upon the type of seal
and its application. About one-half day and two
men are required for a seal change. Some of the
seals are quite expensive, and failure of seals
several times each month was not acceptable.
in
Jiave resulted in be.t£er serving fnr t- h «a f; i±^_njF_r.r>g
.Angeles.,.,,
COORDINATED DESIGN, CONSTRUCTION AND START-UP:
Unless there is active and meaningful interaction
among the designer, construction contractor,
municipal owner and operations personnel, many
avoidable time-consuming and costly problems will
be encountered.
This has been especially true for the City of Los
Angeles wher_e _thjejjt: _ probl gpis h^ye also beejn
ejcacerba t.e.d._— by— — a - ve ry -- s-h-ox.t-__cjpjur_t::manda ted
time-frame for__d,ejs_ign_ and , ^construct ion
-------
fac i LLtx/ _ §L_very complex interdependent expansio n
of the wastewater treatment facility including not
only the state of the art C-G system, but also
state of the art systems for combusting the sludge,
scrubbing the exhaust gases, and cogeneration of
power and steam from sludge combustion as well as
burning of gas from the sludge digestor.
The use of models and pilot testing for improving the design
of a C-G system is often overlooked. Benefits of a model
include a mechanism for checking the design, for ensuring
that all the pieces will fit together and into the allotted
space, for assisting the contractor in construction, and
helping with operator training. Burlington saved $50,000 in
construction costs and undoubtedly saved much more in
change-order avoidance by paying $28,000 for construction of
their model. The limited pilot testing by LACSD helped them
anticipate many problems. Many of the problems encountered
by the City of Los Angeles (discussed in this report) could
have been anticipated and resolved prior to construction by
use of a model instead of requiring considerable extra time
and cost (millions of dollars) for modification of the system
during start-up. I_t is important to note_ th_a_t the
(not met because of ths[~encountere d
p rob 1 em s )__f^r__b^q_uiiii_n q operations was a.l so a sign i f i c a n t
factor contributing to the problems at the City of Los
Angeles' system.
Redundancy should be considered for any component prone to
failure or for which failure could pose a safety risk.
Specifications for equipment should be carefully written for
future systems. Operational experiencing ar-gnir^r) with t-hg
ejuLsLLLDJI system will provide the data needed to prepar e
bet t ejr_s pe c i f i c a 1 1 o n s .
The C-G process is, in effect, a petrochemical type plant and
not a conventional wastewater type treatment process. Safe
and efficient operation of the C-G process requires qualified
operators. Adequate training must be provided. In most
situations to date the operator training may not have been
adequate.
It is essential that any municipal wastewater treatment
authority considering the C-G technology review the
experience and performance of the four municipal C-G systems
discussed in this report. Visits and detailed discussions
with as many of these four system owners as possible should
be conducted so that the experience gained from these systems
can then be incorporated into new designs. Pilot testing of
! untried equipment and modifications to overcome solutions to
i past problems is vital. Accurate operation and maintenance
XI
-------
cost data should also be available for use in estimating
actual costs for future systems.
XII
-------
SUMMARY REPORT
WORKSHOP ON THE CARVER-GREENFIELD SLUDGE DRYING TECHNOLOGY:
PROBLEMS AND SOLUTIONS
Held in Los Angeles, CA
on March 10 and 11, 1987
SECTION 1
OVERVIEW
1.1 Introduction
Four municipal wastewater treatment authorities have selected
the Carver-Greenfield light oil sludge drying (C-G) system
for dewatering their sewage sludge. Ocean County, New
Jersey; the Mercer County Utility Authority (Trenton, New
jersey); the City of Los Angeles, California; and the Los
Angeles, California County Sanitation Districts are currently
designing or constructing C-C systems. The City of Los
Angeles will soon begin operation of their 265 dry ton per
day C-G system. This system will be the first municipal
wastewater sludge C-G unit utilizing the light oil system to
begin operation. Limited start-up tests have been conducted,
and as with any new application of a technology, problems
have been experienced in the start-up of the system. The
problems encountered by the City of Los Angeles have been
aggravated because of the very short time period available
for design and construction and because this system is the
first full-scale municipal system to be built.
To minimize start-up problems at the remaining three systems,
the U. S. Environmental Protection Agency's Municipal
Facilities Division (EPA-MFD) sponsored a two-day seminar and
workshop on the C-G process. Representatives from each of
the four wastewater treatment plants, the C-G design firm
(Foster-Wheeler USA Corporation), the patent holder of the
system (Dehydro-Tech), EPA, the state environmental agencies,
and municipalities currently considering the C-G process
attended the seminar. An attendance list is presented as
Appendix A.
A considerable amount of information on the C-G process was
disseminated at the workshop. Suggestions for improving the
design and construction procedures as well as the EPA/state
funding mechanisms were discussed. The objective of this
document is to summarize the key points of the seminar and
other follow-up information to assist both the attendees and
other individuals who may be considering the C-G process.
1-1
-------
1.2 Report Format
The workshop summary has been divided into three major
sections. Section 2 summarizes the C-G process for those
unfamiliar with this process. In Section 3, a brief
discussion of the operating experiences at several C-G
wastewater treatment plants is presented. A summary of the
recommendations presented at the seminar is presented in
Section 4.
1-2
-------
SECTION 2
THE CARVER-GREENFIELD SLUDGE DRYING PROCESS
2.1 Background
;ln the mid 1940s, Mr. Charles Greenfield began investigating
'processes for separating vitamin oils from ground fish liver
water slurries for his employer, Vitol Processing Co., in New
Jersey. Centrifuging, pH control, and conventional
thickening operations all proved ineffective for drying and
recoverying the resulting vitamin oils and solids. However,
using the fish oil itself as a carrier medium in an
evaporation system was found to be an effective drying
process. When synthetic vitamins displaced natural vitamin
oils, the Vital Company processing plant was shut down. Mr.
Greenfield then set up his own pilot plant and laboratory to
study the drying of edible rendering products and the
production of dry milk powder. This work brought Mr.
Greenfield into contact with Fred. S. Carver, Inc. and in
1955 they entered into a long-term contract. Under this
association, the evaporation procedure for processing the
vitamin oils evolved into what is now know as the
Carver-Greenfield (C-G) drying process.
Development work continued, and in 1961 the first C-C plant
was built in Pennsylvania for processing inedible rendering
plant wastes (waste fats and bones from the meat processing
industry). Initial problems with this system included
non-uniform flow and poor distribution through the 1-inch
diameter evaporator tubes. These problems were mitigated by
increasing the circulation rate of oil slurry in the
evaporator, maintaining the oil-to-solids ratio on a water
free basis at about 5.5 to 1 and devising an improved heater
slurry distribution system. In this plant, the fluidizing
agent was tallow.
In 1964, a C-G plant was constructed for the Hershey
Corporation to dry solids from their wastewater system. Both
primary and secondary trickling filter sludges were dried in
this system. Problems with this system included undesired
thickening of the oil because of the formation of soaps from
fatty acids in the sludges. This formation was intensified
at alkaline pH values and elevated temperatures. This
problem was lessened, but not totally overcome, by
controlling the pH and by using oils, in which the soaps were
more soluble for fluidizing and carrying the sludges through
the evaporators. A heavy oil was used as the fluidizing
agent. An expeller press, used after evaporation and
centrifugation to remove the heavy carrier oil, never worked
and was abandoned. Therefore, the centrifuged product, which
contained 40 percent fluidizing heavy oil, was burned. This
2-1
-------
was not a problem because the original fat content of the
sludge was 25 percent and the added heavy oil used for
fluidizing was readily available at a low price during that
period. The Hershey plant was operated with relative success
until it was shut down in 1974.
Additional C-G plants were built after 1964 in the food,
pharmaceutical, and waste treatment industries. Several
Japanese municipalities purchased the C-G process for
wastewater treatment sludge drying. All of these systems
utilized a "heavy" (non-volatile) oil for fluidizing the
product to be dried. Often the carrier oil itself was
generated from the product being dried. Problems encountered
included plugging of evaporator tubes because of high fiber
content of the sludges and a high oil usage due to the
inability to recover all of the oil from the dewatered
sludge. Increasing the oil-to-dry solids ratio and grinding
of the sludge prior to the C-G process alleviated plugging of
the evaporator tubes. Use of "light" (volatile) oils, which
can be more readily recovered by vaporization, then was
developed as a solution to high oil usage in the C-G*
process. An example of a light oil that might be used is
Number 2 fuel oil, Exxon's Isopar* or Union Oil's Amsco* as
compared with Number 6 fuel oil. The light oil has a lower
viscosity, better heat transfer, is easier to recover and
recycle back into the process, and does a better job of
removing and allowing for the recovery of sewage oil than
does the heavy oil.
The city of Los Angeles Hyperion Wastewater Treatment Plant
is the first municipal wastewater treatment plant with a
light oil C-G system in the United States. Complete start-up
of the Hyperion C-G system is anticipated shortly. Four
other light oil C-G systems are in design, construction, or
operation in the United States (Table 2-1) . A general
description of the C-G process is presented below.
2.2 Multiple Effect System
2.2.1 Process Description
A simplified process flow diagram is presented in Figure 2-1.
This represents the process used in all but the Ocean County,
NJ system. The Ocean County system is discussed in Section
2.3.
*Trade names are given solely for the use of the reader and
do not imply endorsement by the U. S. Environmental
Protection Agency.
2-2
-------
TABLE 2-1
CARVER-GREENFIELD LIGHT-OIL PLANTS
Location
Feed Material
Design Capacity
(Tons Dry
Sol ids/Day)
Disposition of
Dry Solids
Phase
i
U)
City of Los Angeles
Los Angeles, Calif.
Los Angeles, County
Sanitation Districts
Los Angeles, CA
Ocean County Utility
Authority, New Jersey
Mercer County
Utility Authority
Trenton, NJ
Burlington Industries
Clarksville, VA.
F. W. Martinez, Inc.
Concord, CA
Primary/secondary 265.0
sanitary sewage
sludge
Primary/secondary 240.0
sanitary sewage
sludge
Pr imary/secondary 50.0
sanitary sewage
sludge
Primary sanitary 118.0
sewage sludge
Biological sludge 20.4
scouring wastewater
from wool washing
Alum sludge from 16.0
Contra Costa Water
District
Burned in boiler
to generate
electricity
Burned in boiler
to generate
electricity
Fertilizer
Fertilizer
Solids-landfill;
lanolin-burned
and/or sold
Solids-landfill
Construction
Start-up beginning
1/87
Construction
Engineering
Completed
Construction
Start-up
1/88
Operating,
1983-present
Constructed,
Start-up beginning
8/87
-------
The ERM Group.
VACUUU SYSTEM
COOLING
WATER
ADOBACK OF
BOX SOUDS ,
OIL (A
DIRECTION Of
SOUDS FLOW
DRECTION OF
HEAT FLOW
WASTEWATER
VAPOR
N)
\f-3Kt
V.
WASTEWATER
WASTEWATER
EFFECT
RUIOIBNO •
TANK
STEAM
CONOENSATE
SECOND
[VAPORATOR
STAGE
VAPOR
CONDENSATE
THIRD
EVAPORATOR
STAGE
VAPOR
STEAM
CONDENSATE
FOURTH
EVAPORATOR
STAGE
STEAM
CONDENSATE
SOILDS FOR FUEL
OR FERTILIZER
DRY OEOILED
BIOMASS •
9SX SOUDS
OEOILER
CENTRIFUGE
OIL-WATER
SEPARATOR
(SEE FIG.
*>
1^~&
WASTE
WATER
TO POTW
(TO OIL-WATER ^—£\ ^ ^\ ^ £l *~Y
SEPARATOR) C±J CXJ Cr3 £_
S£WAGE « J^^nJfu0"-
/ OIL * WATER) CtNTRATE
CARRIER OIL OIL DISTILLATION _ CARRIER AND SEWAGE OL
-c(T
©
SYSTEM l
(ALSO CALLED A
HYDROEX TRACTOR)
• SOUDS FOR
ADD8ACK (60%
SOUDS IN OJU
SEE F)GUR£
a-4)
(SEE FIG. 2-3)
CRITICAL AREAS; A - AODBACK OF SOUDS TO PREVENT GUMMY PHASE; B - ABRASION AND LEAKING OF PUMP SEALS:
C - ODORS IN VAPORS; 0 - PRESENCE OF FINES IN CENTRATE; E - ABRASION OF CENTRIFUGE-
F - CARRYOVER OF FINES INTO VAPOR SYSTEM. wuraruuc,
FIGURE 2-1. GENERAL PROCESS FLOW DIAGRAM OF THE MULTIPLE EFFECT C-G PROCESS.
-------
Wet sludge from thickening after digestion is fed into a
fluidizing tank. The purpose of the fluidizing tank is to
mix the sludge and carrier oil into a slurry for further
processing. QjLL_and previously dried sol ids called addback
are re-mixed with the wet sludge in the fluidizinq tank. The
additional solids are added to prevent formation of a "cjummy
phase" in the evaporators (further discussion of the gummy
phase is presented in Section 2.2.2). The oil-water-solids
mixture is pumped from the fluidizing tank into the
four-effect evaporator system.
In a multiple effect forced-circulation evaporating system
(Figure 2-1)/ high pressure steam enters the steam jacket
(shell) of the first effect and evaporates water from the
oil-water-solids mixture jUJJ3uo_r_) inside the first effect
evaporator. The incoming high pressure steam condenses in
the steam jacket in the process of heating up the liquor, and
this condensed water is drained out and returned to the
boiler of the system. The water evaporated from the liquor
in the first effect, moves as a vapor (steam, at a lower
temperature and pressure than in the incoming high pressure
steam to the first effect) to the steam jacket of the second
effect. The vapor from the first effect, in turn, condenses
in the steam jacket of the second effect in the process of
heating up the liquor in the second effect, and this
condensed water is drained out and returned to the boiler
system. These processes are repeated in the Jbh i r d _ aad__fo,.urth
eff_ec_t evaporators with the exception that the wajber_
vap_or_l2_ed™"f:rom the liquor in the fnnrth p.££act is—caoideKsed
i,n_a water-cooled external heat exchangers-prior to return to
_the bo iler. The evaporating temperature and pressure is
lower in each succeeding effect.
The liquor (mixture) to be evaporated could move through a
multiple effect evaporating system in either the same
(forward feed) or opposite (backward feed) directions as the
vapor is moving. Forward feed results in a simpler
installation because liquor can be transferred between
effects without pumping. However, it is seldom used if the
liquor becomes more viscous as it evaporates. Backward feed
is then preferred because the most co.ncen-t-r-a~fe@d—iiq-uo-r—is
evajjgrajted in the first effect where_ <~h^>—high—t_empj2jia_ture
reduce s vj.sc o s i ty__ an d jm i_n imi^ze s_ _Lts__ad-V-exs.e affect ,,Qjg__he a t
JLr a n s-f SLT_ c o e f f Tc i e n t s . In the Carver-Greenfield system,
backwardfeed is use.
When backward feed is used in a multiple effect system, a
unique numbering system is used to identify the vapor and
liquor streams. The term "stage" is used to refer to the
sequential order in which the liquor (the sludge-oil-water
mixture) moves through the system. Thus, in a four-effect
(four evaporator) system, wet sludge liquor enters the first
2-5
-------
stage (last effect — 1 owes t temperature and pressure
evaporator). This liquor sequentially flows through the
second and third stage before treatment in and exiting from
the fourth stage /first- effect—highest temperature and
evaporator! . On the other hand, the previously
described effects (evaporators) are numbered in the direction
of heat flow. In the C-G process, the fresh (highest
temperature and pressure) steam enters the first effect
evaporator shell which is the fourth stage for liquor
evaporation (where the solid's content is the greatest and
hence the boiling point of the liquor is the greatest), jieat
then s_u_b..s.e.qu.enj:i.ally flows as vapor through the second ,
th_ir_d_, a n d__fjpja_r_t ii __ef_f e c t s .
A multiple effect evaporator system may appear to present a
difficult control problem due to the number of evaporators to
control (systems of six evaporators are common), but in
actuality, control of rnul_ti pie, effect systems JLs_jifiJj*.l:j_y_gJ.y
§jjapJ-e.. Liquid levels have to be maintained in each
evaporator, &£e_a,m_ must be supplied to the first effect at
and water must be supplied to the
water-cooled external condenser at a reasonably constant
temperature and adequate flow rate to condense the vapor from
the last effect. Th e 1 i quo r .__j eed ____ _£Lo.w__ r.a t e . __ i.js___ttjen
established to _ produce the _______ d,e_si_red-__^jjiaj____^oj.-ija s
concentration. When t h i s Ti 3orfe7~"aii of the intermediate
vapor temperatures and liquor concentrations establish
themselves and are not subject. to external control.
A phenomenon that reduces the potential value of the
available temperature drop between the condensing steam
entering the first effect and the water cooling the vapor
from the fourth effect is the boildng point rise (BPR) that
occurs when a liquid that contains dissolved salts is
concentrated. For example, a 22 percent aqueous solution of
sodium chloride boils at atmospheric pressure at a
temperature of 222°F, whereas pure water boils at 212 F
(therefore, the BPR is 10°F.) The vapor evaporating from the
aqueous sodium chloride is pure water and will condense at
212 F at atmospheric pressure. Thus, some of the available
temperature difference disappears. There are more losses in
the available temperature difference when the liquor retains
some superheat after it passes through the evaporator tubes.
This BPR phenomenon reduces the amount of energy per hour
that can be exchanged in the system (capacity of the system),
but does not lower the steam economy (kilograms [kg] of water
evaporated from the liquor divided by kg of steam charged to
the first effect) . Approximately one kg of water is
evaporated from the liquor for each kg of vapor condensed in
each effect (evaporator) , assuming minimal heat loss from the
system and adequate temperatures and pressures in each
2-6
-------
effect. The approximate steam economy is calculated by
dividing the steam used by a single effect evaporator by the
number of effects. In general, the multiple effect systems
are designed based upon the amount of water to be evaporated.
An important advantage of a multiple effect system is that
relatively high rates .of _ evaporation per __ unit pf'JKeat input
qajLJie^ai^Le.yed . For example, an evaporation to steam ratio
of about three can be achieved with a four effect system.
in the C-G system, not only water, but also some of the
carrier oil and sewage oils in the sludge will be vaporized
from the sludge in each effect. Condensation of this steam
and oil occurs in the heat exchangers shell section of the
evaporators. The condensate i s . directed _to -An— rv--i-V— ua-h.^r
separator which is usej3_t-Q_-S-epa-c-ate — water — an-d — recover- — the
jjjUL. .Th.e _r e coyer ed_oJJ__i-s — Recycled — to — the — fJoiijiizJJig_tajnk .
^ t . The
_
location of the oil separation system with respect to the C-G
process is shown of Figure 2-1. A more detailed schematic of
the water-oil separation system is provided in Figure 2-2.
After evaporation, the solids-oil slurry contains
approximately liz^.Q.JSeT^en^^solj.o^ in oil,, Water content is
typically on the order of L^2_ pe r c eritT~~Th e terminology used
to refer to solids concentrations varies depending upon the
location of the solids. The solids concentration of the
sludge entering the first-stage evaporator is measured with
respect to the water content. Therefore, a 10 percent solids
concentration fed to the evaporators represents 10 percent
solids in water. Oil is present at a ratio of roughly eight
parts oil per one part dry solids; however, the oil content
is not included in expressing the solids concentration prior
to exiting the evaporators. Conversel v_^._o_n.C-e _ t.h.e._s 1 udq e
e x i t s __t h e_e_va pjorat, pj^s..*_ ,th.e_l 5^-2 0_.p e,r c e n t__solj. d s _c_o n c e n t r a.t ion
r e far s^fco~the-H3e-r^e^tag-e_oJL_s.o_l_ids in oil and Jjeg 1 e c t s „ .t h e
approximately 1-3 percent water content.
Approximately one-half of the dewatered sludge flow leaving
the last stage (first effect evaporator) is recycled to the
fluidizing tank in a process called addback (see Section
2.2.2) . The oil present in the remaining dewatered sludge
must be removed from the solids so that the oil may be
reused. Bulk separation is accomplished by ce^n^rjjEuaaJt.ipn to
increase the solids concentration to approximately j[0__pe_r_cent
(solids in oil, Figure 2-1). Centrate from the centrifuges
is sent to a fJLasti_js±jJLl . This still separates the sewage
qil_from the carrier, oil, thereby allowing the carrier oil to
~ A diagram of the oil distillation system is
___
presented in Figure 2-3.
The ability to recover the 10-20 percfent content of sewage
oil normally present is sewage sludge is important for at
2-7
-------
The ERM Group.
OILY WATER
FROM ALL
EVAPORATORS
FRESH
CARRIER OIL
CARRIER OIL TO
FLUIOIZING.
EVAPORATION,
FLUSH SYSTEMS
OIL FROM
COALESCER "B"
WATER FROM
COALESCER "B"
OILY WATER
SURGE TANK
PLATE
TYPE
SEPAR
ATOR
OILY H20 TO
COALESCER ,
PREFILTER COALESCER
RECOVERED
OIL PUMP
TO TANK "B
FRESH/RECOVERED
OIL SURGE TANK
DEOILED
WATER TO
OUTFALL
FIGURE 2-2.
SCHEMATIC DIAGRAM OF THE OIL RECOVERY SYSTEM FOR THE
C-G PROCESS.
- -V" 1 rnifcii .
TtMERN Group,
-------
The ERM Group.
OIL/WATER VAPOR
TO STAGE 1
EVAPORATOR
HEATERS
OIL VAPOR
TO STAGE 4
EVAPORATOR
HEATERS
CENTRATE OIL
FRO)
CENTRIFUGES
VENT VAPORS
VAPOR FROM FIRST STAGE
r HYDROEXTRACTORS
VAPOR FROM FLASH
STILL "B"
CENTRATE
SURGE TANK
- aASH STILL
FEED PUMP
185 PSI STEAM
STEAM
CONDENSATE
TO RECEIVER
FIGURE 2-3.
FLASH
STILL
t
- TO VENT GASES CLEANING SYSTEM
-i VAPOR FROM SECOND STAGE HYDROEXTRACTOR
-V VAPOR FROM STRIPPER "B"
/
OIL
STRIPPER
V/L
SEPARATOR
FLASH STILL
CIRCULATION
PUMP
SEWAGE OIL
TO STORAGE
SEWAGE OIL
DISCHARGE
PUMP
-?
STRIPPING
STEAM
SCHEMATIC DIAGRAM OF THE OIL DISTILLATION SYSTEM
FOR THE C-G PROCESS.
-------
least two reasons. First, if this oil is not separated from
the carrier oil after centr i f uga tion, it could cause the
formation of emuls ions and foaming problems in the C-G
system. Secondly, the sewage J3J.1S can be used to supply heat
ajid_riowe r . For example, when separated from sewage sludge at
the 13 percent sewage oil content, the sewage oils can be
used as a fuel in boilers to generate 100 percent of the
steam needed for producing steam for the C-G system. Khen
separated at the 16 percent sewage oil content from sewage
sludge (such as at Trenton, NJ) , the sewage oil can be used
not only to generate 100 percent of the steam required by the
C-G system, but also to generate 50 percent of the C-G system
power requirements.
Some solids may be present in the Qent£a_£e, and may interfere
with the operation of the f lash__still ; thus, centrate quality
is a potential concern (see Section 4.5.5). Fine particles,
in particular, are difficult to remove by ^ej3jLr.i.f.aga_tlojj . As
the carrier oil is separated from the sewage oil in the flash
still (Figure 2-3) for recycling back into the evaporating
process, the solids in the sewage oil become concentrated.
The resultant solids content in the sewage oil is about ten
times the solids content of the centrate. Too high a solids
content, such as 50-60 percent, is n_o_t _pjjmpable and the
a p . ~~ '
In the heavy oil C-G process, additional oil recovery beyond
centrifugation is usually accomplished using a screw type
press. The oil-solids mixture is either burned, or depending
upon the oil used, is used for animal feed or fertilizer.
The cost of oil for a heavy oil system for a large plant can
be very high. Use of heavy oils at the Los Angeles Hyperion
plant, for example, would have cost approximately $3 million
per year. Light oil systems have therefore been specified
for all four municipal wastewater systems in the United
States. Oil costs for these systems are projected to range
from approximately $10,000 (Ocean County) to $150,000 (LACSD)
per year.
In a light oil system, final oil removal from the centrifuged
sludge takes place in an unit operation called
^xfiydroextract ion, which sounds like a misnomer for the
process. However, since ste_am ___ is utiLized _ to _vapojLJz.e_ the
oil and thereby remove the light oil from the solids, the
terminology was utilized. By the time the solids reach the
hydroextractor, nearly all water has been removed. Oil, not
water, is then vaporized from the solids. Thus, the term
"de-oiler" may also be appropriately used as well as the terra
"hydroextractor" .
The hydroextractor is essentially a two-stage dryer. Oily
solids enter a heated vessel maintained under vacuum where
2-10
-------
hollow steam-heated paddles are used to mix and convey the
solids. Any residual moisture and the carrier oil are
vaporized thereby drying de-oiling the sludge. Although a
hydroextractor somewhat resembles a screw conveyor, the
solids actually flow by gravity from the first and second
stage of the hydroextractor. Approximately 95 percent of the
oil is removed in the first "stage, with the remaining oil
being removed in the second stage. The vaporized oil is then
recondensed in the evaporator system and reused. The final
dry solids product exits the hydroextractor.
Dust carry-over into vapor lines is a potential problem with
the hydroextractor, particularly if fine particles are
present, if large amounts of steam are used to reduce the oil
content to a small amount, or if the influent solids
concentration is low (<50 percent) . At low solids
concentration, the quantity of oil evaporated is high,
creating a high vapor velocity. The high velocity through
the hydroextractor increases the generation of dust. A
ba f f 1 e . i.s_.u^jed_t^_heJ.B_^L]^e^ejit—so. 1 ,ids_aarr-V^a&er .
Odors are a potential problem with any sludge drying
operation including the C-G process. Hot wastewater
treatment sludge has the potential to be very odoriferous.
To minimize the potential for odors, as well as fires, all
process units in the C-G system are _seaJLe<3. Gases from the
process units are vented into vapoj:__l.ines. Oil in the gases
is recovered from the condensate of the heat exchanger. The
remaining gases must then be scrubbed in the evaporator or
are burned in the boiler to remove noxious odors. Activated
carbon could also be used for odor removal, however, the cost
for such a system would be expensive. Therefore, the vented
gases are typically burned in a C-G application in the boiler
that produces the steam for. the system.
2.2.2 Addback
Addback (Figure 2-4) is the process which dewatered
solids-oil slurry from the fourth stage (first effect)
evaporator is recycled to mix with fresh sludge in the
fluidizing tank. The purpose of thijs_pj-0£ess__i_s J:o_pr_e_v.snt
formation of a "g,nmm.y phase" which can be a significant
operating constraint.
The fluid properties of the solids-water-oil slurry change
depending upon the solids to water concentration. Wet sludge
from the wastewater treatment system typically contains from
2 to 20 percent solids depending upon the prior dewatering
steps (e.g., belt press or centrifuge). Solids
concentrations (water basis) mixed in the solids-water-oil
slurry in this range are pumpable. As the solids
concentration increases above 20 percent (water basis) the
2-11
-------
The ERM Group.
WASTE-
ER
DIRECTION Of
SOUDS FLOW
DRECTION OF
HEAT FLOW
i
i—>
N)
FE£D
FLUIOIZJNO
TANK
4TH
EFFECT
WASTEWATER
WASTEWATER
WASTEWATER
FIRST
VAPORATOfi
STAGE
•ADOBACK SOUDS AT
0U SOUDS IN WATER
(eox SOUDS IN OIL)
DRY DEOILED
BIOMASS
PRODUCT
(88X SOUDS)
CARRIER OIL
TO RECOVERY
SYSTEM
HOT DRY
BIOMASS AT
60X SOUDS
IN OIL
EXCHANGER
HEAT
EXCHANGER
EXCHANGER
CRITICAL AREAS (NOT SHOWN PREVIOUSLY): A - PLUGGING OF SPIRAL HEAT EXCHANGER
B - VISCOMETER AND RATIO CONTROLLER
FOR ADDBACK CONTROL
NOTE:
AOOBACK
APPROXIMATELY 60 PERCENT
OF THE SOUDS ARE RECYCLED
AS ADDBACK TO AVOID
DEVELOPMENT OF A 'GUMMY
PHASE* IN THE EVAPORATORS
L
FIGURE 2-4. ADDBACK SOLIDS FLOW DIAGRAM FOR THE MULTIPLE EFFECT C-G PROCESS.
-------
solids begin to coalesce and become sticky (hence the term
"gummy") . This sticky or gummy phase occurs in the range of
ZQ to 30 percent solids (water basis). The viscosity of the
"gummy" slurry increases dramatically, and the slurry becomes
difficult to pump. Once the solids concentration increases
past approximately 30 to 35 percent (water basis) mixed in
the solids-water-oil slurry, the solids again are dispersed
in the slurry and are pumpable.
Should the gummy phase occur within any of the process units,
plugging of the unit would result. The unit would then have
to be shut down—f^r__cie_an_ing. Dry solids and oil from the
fourth-stage evaporator are thus added back to the wet feed
sludge to increase the solids content of the slurry to
approximately 35_pejccent solids in water (Figure 2-4).
The slurry of dry solids in oil from the last stage
evaporator is very hot (e.g., 250 F). If this slurry were
added directly to the fluidizing tank, vaporization of water
and oil would occur in this tank. Considerable heat value
would also be lost. Instead of recycling the addback solids
directly to the fluidizing tank, the solids pass through a
series of spiral heat exchangers (Figure 2-4) . The spiral
heat exchangers are constructed of two parallel stainless
steel plates which are rolled into a spiral cylinder.
Spacers are used to separate the plates during rolling to
form two parallel but separate flow paths. Sludge from each
effect is fed into one flow path of its corresponding spiral
heat exchange while the addback solids are pumped through the
other path. Th_e_axldbacJc__s^_ljLcLs._a,rje__u,s.ed t_o—pJLeJi.ea.t—Lhe
sludge JLed. to.?Teac"h"~e"f£ectljH±-h.erj5.by__J.inProving the ener gy
efficj1ejrcy__gf the proc_esis,.
Two of the problems with addback are start-up and jp_r.g_cej3s
control. During start-up of a new plant, dry solids will not
have been generated to feed back into the fluidizing tank.
Compost or some other source of dry solids will thus be
required. Care should be taken to ensure that the compost or
other material is free of particles which may clog the
system. Secondly, care must be taken in c_Qn.tjc_o,lling__the_ra_te
of addback which can be difficult. Typically, about half of
the dried solids from the last stage (first effect
evaporator) are used as addback (i.e. a solids to recycle
ratio of 1 to 1) . Y.iscomete.r_s__on_the discharge_p.ump.-£roitu±Jije
f luidj.jz.ing.._ tank are" now recommended to control__addjb.aj;k. If
the viscosity increases (thereby indicating the incipient
development of the gummy phase), alarms are sounded and
additional dry solids are fed to the fluidizing tank. A
ratio controller is used to physically control the amount of
solids recycled as addback.
2-13
-------
2.3 Mechanical Vapor Recompression
Mechanical Vapor Recompression (MVR) evaporators offer an
alternative to the multiple-effect evaporator system
described above. A full-scale MVR C-G system has been in
operation since 1983 in the Netherlands in a heavy oil
rendering application. The MVR alternative will be used by
Ocean County, New Jersey, which will represent the first
light oil MVR evaporation system for drying municipal
wastewater sludge.
In an MVR system, water vapor generated by steam in the first
stage evaporator is compressed and recirculated to the steam
chest oQf first-stage evaporator at a temperature of about
170-180 C and pressure of about 20 psia (Figure 2-5) . The
compressed water vapor condenses and is removed providing
enough heat to vaporize an equal amount of water, thereby
concentrating the sludge in the evaporator.
The differential in temperature between the compressed water
vapor and the liquor is about 15 F. The magnitude of the
temperature difference between the condensed steam and the
evaporating water in the MVR system is of prime importance.
The lower the temperature differential, the lower the
electrical energy required f°ro recompression. At a
temperature differential of 150 F, about 6 to 7 pounds
water per pound of steam equivalent can be removed in the MVR
system compared with about 3 Ibs/water/lb/steam in the
four-effect evaporating system. These calculations hold
provided that boiling point rise (BPR) (previous^ discussed
in Section 2.2) is not higher than about 4 or 5 C. If the
BPR is not excessive, the final - temperature differential at
the exit of the vaporizer can be low and the evaporating
capacity of the system will be high. With this low
temperature differential, only a moderate amount of
recompression is needed to attain the necessary pressure and
temperatures of the saturated vapor. At the Ocean County,
New Jersey, facility, a 950 HP (650 KWH) compressor will be
used for the MVR system in which there is also a two-effect
forced circulation evaporating system in series after the MVR
unit for final drying.
Influent sludge solids contents of not more than about 8
percent have been felt important for the Ocean County system
largely because of the ease thereby in avoiding the "gummy
phase". Addback of partially dried solids can be avoided by
doing the bulk of the evaporating in the MVR unit, provided
the feed sludge to be dried is properly metered into the
sludge-water-oil slurry (liquor) in the MVR unit (e.g., 20
parts liquor at 50 percent solids to 1 part feed at 7 percent
solids). In this manner, the percent solids (water-solids
basis) is then sufficiently high that the critical
2-14
-------
The ERM Group
FEED
BIOMASS
O 5-15X
SOLKT
FLUID TANK
•VENT
VACUUM SYSTEM
CARRIER
OIL
MECHANICAL
VAPOR
RECOM-
PRESSION
EVAPORATOR 20
WASTEWATER
VAPOR
2ND
EFFECT
WASTEWATER
s&~
FIRST
:VAPORATOR
STAGES
VAPOR
1ST
EFFECT
SOLIDS FOR FUEL
OR FERTILIZER
SECOND
i VAPORATOR
STAGES
^TC^ J c
STEAM
OIL-WATER
SEPARATOR
(SEE FIG. 2-2)
WASTE
WATER
TO POTW
_ STEAM "-—,—-^ I STEAM
CONDENSATE J CONDENSATE
(TO OIL-WATER V_/TY (TO OIL-WATER
SEPARATOR) £±_} SEPARATOR)
BY-PRODUCT __ DRY
" OIL + OIL
DRY BIOMASS ^NTRjfUGE
I MIAI- |—
CARRIER
OIL
OIL DISTILLATION
SYSTEM
CARRIER OIL
DRY DEOILED
BIOMASS O
98% SOLIDS
(ALSO CALLED A
HYDROEXTRACTOR)
SEE FIG. 2-3 FOR DETAILS
FIGURE. 2-5. PROCESS FLOW DIAGRAM OF THE MVR/C-G SYSTEM.
-------
water-solids ratio is never encountered. Besides eliminating
addbacks, considerable economies can also be realized using
the MVR unit because only simplified equipment is needed for
dewatering the sludge prior to C-G processing. Boiling point
rise is not likely to be a problem with sludges in MVR
systems until the solids contents reach about 60 percent
(water basis).
Operating experiences with a light oil MVR C-G system are not
yet available. Thus, it is not yet know how the economics of
an MVR system will compare with a four-effect system as well
as with other alternatives for drying wastewater treatment
sludges.
2.4 Light-Oil Systems
A summary of the C-G light-oil systems in operation, design,
or construction are presented in Table 2-1. The four
municipal wastewater treatment authorities: Los Angeles City;
Los Angeles County Sanitation Districts; Mercer County, New
Jersey; and Ocean County, Mew Jersey are the only full-scale
municipal wastewater sewage sludge drying systems in design,
construction, or operation in the USA. Table 2-2 provides a
summary of the estimated capital and O&M costs for these four
systems and the Burlington Industries system. AJ..1 Q_fL_..the
municipal systems were funded as IjmQj£a_iJ^Le_, under the
Innovative/Alternative (I/A) provisions of the Clean Water
Act.
Dehydro-Tech Corporation through Hanover Research Corporation
owns the patents for the C-G process and licensing fees are
negotiable. The estimated capital and bid costs of the
Hyperion C-G System were respectively $43.2 and $30 million,
of which the license fee will range from $1.4 to $1.7
million, depending upon.jthe._ abil ijty _to meet performance
guarantees. The actual cost as of March 1987, was $43
million (Table 2-2) . The influent solids concentratfltmrs-j
projected energy and oil requirements, and the operating
schedule for the five U. S. C-G systems is presented in Table
2-3. The two Los Angeles systems will operate continuously.
It is anticipated that redundancy in the system (e.g., extra
pumps, centrifuges, etc.; Table 2-4) will permit routine and
required maintenance to be performed without adversely
affecting these systems. Operational experience will
determine if sufficient redundancy has been provided.
2-16
-------
TABLE 2-2 SUMMARY OF UNITED STATES C-G LIGHT OIL SYSTEM CAPITOL AND OiM COSTS (IN MILLION
System
City of Los Angeles,
CA. Hyperion Plant
Los Angeles County
Sanitation Districts
Mercer County, NJ
Trenton, NJ
to Ocean Co. Mastewater
1 Authority, NJ
Burlington Industries
Clarksville, VA
Descr iption
4-effect C-G system
4-effect C-G system
4-effect C-G system
MVR 2-effect C-G system
5-effect C-G system
Capital Costa
Carver-Greenfield Total Project
Est. Bid Spent(a) Est. Bid
43.2 30.7 43 224 158
60 54.7 (c) 48.7 120
12 12 35
38 13
3.3 4.23
OtM Costs
Less(b)
OiM Benefit
(5/ton) (5/ton)
40-50
80 7
N/A N/A
N/A N/A
30
Total Costs
(9/ton)
160-200
248
N/A
N/A
-------
TABLE 2-3
SUMMARY OF UNITED STATES C-C LIGHT OIL SYSTEMS
OPERATIONAL INFORMATION AND ENERGY REQUIREMENTS
NJ
I
oa
System
City o£ Los Angeles, CA
Los Angeles County
Sanitation District, CA
Mercer County, NJ
Ocean County Wastewater
Authority, NJ
Burlington Industries
Clarksville, VA
Description
4-e££ect C-C
4-effect C-G
4-ef£ect C-G
MVR 2-ef£ect
5-e£fect C-G
system
system
system
system
system
Operational % Solids % Solids Projected Energy Requirements
Days per in water* in oil BTU/LB Make-up Oil**
week in feed in feed water gallons/day
evaporated
7 20 11 363 174
7 19 9 362 264
(2 trains)
5 22 17 363 110
3 7 10 200 44
5 4
« Composition of outgoing sludges 2.5-3% water, 0.1 to 0.2% Amsco oil, 1.3 to 2.3% sewage oil, 95% solids.
** Cost per gallon is 51.50 as of March 1987.
-------
TABLE 2-4
SUMMARY OF UNITED STATES C-G LIGHT OIL SYSTEMS-SYSTEM REDUNDANCY
City of LA,CA LA County San. Dist.,CA Mercer Co.,NJ
On Stream Time
Number of Trains
Fluidization
Evaporation
Oil Removal
1° Oil Distillation
Oil Recovery
Dilute Acid
Vent Gas
Ni trogen
Start-up Boiler
Cool ing Tower
Control Approach
Continuous
2 + 1 (a)
1 + 1
2 + 1
2 + 1
1 + 1
1 + 1
1
1 + 1
1
Temporary
None
Semi-Automated
Operation
Continuous
3 +
1 +
3 +
3 +
1 +
1 +
1 +
1
1 +
1
1
Spares
1
Spares
Spares
1
1
1
1
Fully Automated
Operation
5 Days/Week
1
1 + 1
1
1 + 1
1
1
1
1
1
1
W S A C
Semi -Automated
Operation
Ocean Co., NJ
3 Days/Week
1
1 + 1
1
1
1
1
1
1
Bottles
1
1
Semi-Automated
Operation
WSAC
Wet Surface Air Cooler (a) 2+1=2 operational units plus 1 redundant unit.
-------
SECTION 3
OPERATIONAL EXPERIENCES
3.1 Operating Systems
The operating experiences of three wastewater C-G light-oil
systems were discussed during the seminar: 1) the limited
start-up experience of the City of Los Angeles Hyperion
system, 2) the Burlington Industries industrial wastewater
system in Clarksville, Virginia, and 3) a pilot-scale system
operated for 18 months by the Los Angeles County Sanitation
District (LACSD). Brief discussions of some of the operating
experiences at these systems are presented in this section.
More detailed discussions of specific problems and
recommendationsare presentedin Section 4.Discussions of
the operating experiences at other systems can be found in
Reference 1.
3.2 City of Los Angeles Hyperion C-G Treatment System
3.2.1 Project History
The City of Los Angeles Hyperion Wastewater Treatment Plant
serves approximately 3.4 million people in a 640 square mile
area. Approximately 420 million gallons per day of
wastewater are treated at this facility. Since the treatment
plant is subjected to marine secondary discharge standards,
not all of the wastewater currently receives secondary
treatment. Currently, only about 100 MGD receives secondary
treatment. The combined plant effluent of primary and
secondary effluents is then discharged to the Pacific Ocean
through a five-mile outfall. Primary and secondary
anaerobically digested sludge from the treatment plant is
either landfilled or discharged to the ocean through a seven
mile outfall.
In the early 1970s, the city began investigating alternative
means for sludge disposal. State and federal agencies
required that faster progress be made towards developing a
better sludge management system. Legal actions resulted, and
a Consent Decree was signed in late 1979 or early 1980. The
provisions of the Consent Decree specified that design of a
sludge treatment system had to be completed by December 1982.
Many fcerbni^] issues
r e s o lutiogLnf +•*»**«» issues required an additional two yea r s
JLQ resolve! Thus, the New Jersey based engineering firm,
P^h0r^h^rUrr-^rr"n"ot able to begin design and detailed
engineering of the C-G system until February 1982.
3-1
-------
The Hyperion Energy Recovery System (HERS) , of which the C-G
system is a component, is the largest I/A project funded to
date by EPA, and as of March 1987, was 98 percent complete.
The HERS consists of several components: mechanical
dewatering of the digested sludge using centrifuges, the C-G
process, combustion of the sludge for steam and power
generation, and combined-cycle digestor gas power generation
using gas and steam turbines. Figure 3-1 presents a
schematic flow diagram of the Hyperion treatment plant
including HERS.
3.2.2 Cost
Due— t-Q— th.e_JJjgjLted time . _f_rame.-ayailabJ._e for design imposed by
t-hj5 C o n s ej^t— Hec-r-e-e-, — a-d-gg+i-a-t-e — te-J-m-e — wa-s — oo-t— -a-v.ajLI.ab _l.e__f o r
sufficient quality control and checking of the design
drawings. The initial estimates for the constructon of the
entire HERS system by the various design engineers involved
was $222 million. The construction bids, however, were
significantly less, and the project was awarded to the
general contractor at $158 million. The lower bid price was
believed due to bidding during a depressed time for
construction jobs. Due to factors such as design changes fo_r
^ i i-y of the system and the insufficient time
q "*_]_•< <~y ^nnt-mi — off the nriryjnal design drawings, numerous
change orders were needed. Change orders as of February
1987, have totaled $58 million. Thus, the total project
construction cost as of February 1987, has been $216 million
or within $6 million of the original estimate. Under the
construction grants program, however, funding for the project
was limited to the contractor bid of $158 million plus 5
percent for change orders totaling $166 million. Thus, the
city has had to fund an additional $50 million for
construction of the system.
3.2.3 Operational Problems
Limited start-up testing has been conducted at the Hyperion
plant, and several problems, many of which have been
overcome, were encountered during this start-up. For
example, compost was used to provide the solids needed to
begin operation of the system and avoid the gummy phase.
Wood chips present in some of the compost were too large to
pass through the spiral heat exchanger clearance of 5/16 -inch
to 7/16-inch, and the heat exchanger became plugged. The
exchangers were cleaned and the wood chips were subsequently
screened out. r_avitation in pinch valves due to high
^ ffa^aflHai pressure through the valves was also a problem.
•Replacement of these valves was a rather simple but very
costlv solution in both time and money. Flow measurement has
been another problem that is being approached differently by
the various municipalities. The best flow metering solution
3-2
-------
Tfi9 ERM Group.
HYPERION ENERGY RECOVERY SYSTEM
U)
I
PRIMARY
SLUDGE
•—•*»•
REMOVAL
—
COMPRESSION
DEHYDRATION
FILTRATION
r
ANAEROBIC
DIGESTION
I
-H~
MECHANICAL
DEWATERING
.._.
—
GAS
TURBINES
t
ELECTRIC
POV*R
—
-»*•
WET CAKE
STORAGE
—
^
H
•-ii ^^
HEAT
RECOVERY
STEAM
GENERATION
CARVER-
GREEHnELD
DRYING
PROCESS
DISPERSION
STACK
STEAM
TURBINES
ELECTRIC
POWER
SDF
STORAGE
*-
FLUE CAS
CLEAHUP
t
HEAT
RECOVERY
STEAM
GENERATION
I
THERMAL
PROCESSING
A _
WASTE
ACTIVATED
ILUOOC
^
WAS
THICKENING
STEAM FLOW
GAS aow
SLUDGE SOUDS FLOW
FIGURE 3-1
SCHEMATIC OE THE HYPERION SLUDGE
PROCESSING FACILITY.
-------
is not known at this time. Mass flow meters were used at
HERS to monitor solids flow rates at key points.
,Fa^Ture of pump goaig has been a significant problem. Due to
tKe"~~prrs"s i b 1 i ty for potentially dangerous hot hydrocarbon
leaks, double mechanical seals had been specified. The
inside seal exposed to the water-solid-oil slurry usually
failed within 3 to 10 days of installation. The use of
extremely hard tungsten carbide on both faces of the seals
-has apparently resolved the problem — n?;3r ^-o^m" 5
non-flammable oil is used as a gearing and cooling — media
between the double seals. Design engineers have recommended
that the inside seal, exposed to the sludge, be flushed using
oi_l as the flushing agent. _TJa.e — pumps — ^lir rpntlv being
'utilized were designed for more routine wastewater sludge
handling needs. Precise-tolerance chemical-quality pumps are"?
perhaps more appropriate for a C-G system. — '
Another significant problem encountered was smoldering of
srTM rig within a section of the process piping. Because of
the problems encountered during construction, the steam
generation system was not on-line when the C-G process was
undergoing start-up. A temporary skid-mounted boiler system
was thus installed to permit operation of the C-G system.
Partial failure of the temporary process boiler allowed a
section of pipe to cool. Soli/l*? in ^^"3. pipeline then
accumulated on tb*» int-pri-ojs — p>pn \jjjTi i_s_. _Air subsequently
entered the pipeline when _a — coal was inadvertently lef-t out^
.g£— ±£±_ -L£S££ni ^nrir\a_ routine ' maintenance. The air allowed
auto-oxidation_of the fatty acids in the solids on the pipe
walls to occur . This exothermic reaction generated
sufficient heat to oxidize other orqanirsr and smoldering
continued. It is estimated that a maximum temperature of
1300°F had been reached in a six-foot section of pipe by the
time an operator noticed that the pipe insni*i-inn was
saqgjLaa. . Steam flow was then increase^ to cool the pipe
(steam is considerably cooler than 1300 F) . Damage to the
system was thereby limited to the heated section of pipe and
nearby plastic valve seats. Design engineers stressed that
no explosion occurred and that other drying technologies
(e.g., flash dryers, rotary kilns) have a higher risk of fire
or explosion than the C-G technology. After this incident,
this section of the C-G hydroextraction system was redesigned
to prevent air from entering vapor transport lines by raising
'fh^_np^r=iting pr^qB-i^" fmm u^nnm fo_ atmospheric. Nj.tr ngej
blanketing of key processing units during maintenance was
also added to reduce fire hazards.
Pining Changes were also made and were being implemented in
March 1987. Smaller pipe sizes are being used t.a inrrftaae
the flow velocTt7""to"5JL-l^t per second or greater in the
lpor lines! Soot blowers (special cyclonic dust collectors)
3-4
-------
are also being installed to help clean out vapor lines, hhen
the pipe changes are constructed, some loss of flexibility in
operation will result; however, safety will be improved.
3^.3 Burlington Industries
3.3.1 System Description
The Burlington Industries textile plant in Clarksville,
Virginia, investigated pyrolysis, multiple hearth
incineration, and fluidized bed incineration before selecting
a five-effect C-G system to handle wool scouring wash water
and bio-sludge. Their reason for selecting the C-G system
was not justified on the basis of cost per ton for dry
solids. Instead, their system was selected to accomplish the
following objectives:
1. To avoid the lagooning of the malodorous wastewater
from wool scouring by drying the solids. The
resultant dry lanolin from wool scouring can now be
sold and the dried biological solids (biosludge)
from cleaning the finished wool are now acceptable
for disposal at the county landfill.
2. The second objective was to decrease chemical
consumption, cost and quantity of resultant solids
generated at the wastewater treatment plant
because of evaporative drying capability.
The lower cost land application alternative for sludge
handling was prohibited by the Virginia Health Department.
Furthermore, air quality concerns, eliminated alternatives
involving incineration.
Capital expenditures were $4.13 million for a five-effect C-G
system with an original design capacity of 20 tons per day
solids 3nH w^ter evaporation rate of 19_^5.aa_aQJJU3i3s of water
per hour. Currently, the system is required to process only
6.0 dry tons of solids per day. Construct ion began in
November 1981, and completed inCJajiua~ry 19JT3T. A limited
starf-np ™*s Conducted in February L983; however., official
start-up was begun in September 1983. Annual operating costs
are approximately i? I 8U , UUO "adjusted for depreciation and
resource recovery. Recovered lanolin is sold and helps
offset the cost of operation. The net cost of the Burlington
C-G system is approximately $430 per dry ton which includes a
cost allowance for recovered lanolin. As previously noted,
the cost per dry ton of solids produced was not the criterion
Burlington used for choosing the C-G technology. Instead,
Burlington chose the C-G system to solve their difficult
3-5
-------
problem of recovering very soluble solids from dilute a
wastewater (0.5 to 1.0 percent solids).
3.3.2 Operational Problems at Burlington
Process control was initially a problem as changes in the
solids concentrations interfered with the flow meter
operation. Pressure control helped overcome part of this
problem. Major process problems experienced included
inaccurate flow measurement (corrected by replacement of some
flow meters and by manual half-hourly solid-water-oil ratio
determinations) , excessive loss of the carrier oil (to be
corrected by adding a more efficient oil-water separator),
and centrifuge plugging and wear (corrected by flushing the
centrifuge with carrier oil and installing removable wear
plates in the centrifuge). Plugging of the vapor lines with
solids was also a problem. This problem has been practically
eliminated by modifying the piping to increase the velocity
within the vapor lines and manual cleaning of the five-foot
long vapor line between the de-oiler and the scrubber. To
clean out the small amount of solids that build up takes
about five minutes once during each shift. Additional
operational problems and the corrective measures are
summarized in Tables 3-1 and 3-2.
Burlington Industries reports that, in general, their C-G
system is operating satisfactorily and is currently
encountering routine maintenance problems which consist
mostly of leaks and seal failures. Burlington still,
however, is dealing with excessive pump erosion problems.
3.4 LA County Pilot System
3.4.1 System Description
The Los Angeles County Sanitation Districts (LACSD) operated
a two-effect C-G pilot plant system for 18 months in order to
generate sufficient dry sludge for combustion and air quality
testing. The LACSD intends to incinerate the dried sludge
and needed to ensure that the incinerator emissions would not
exceed air quality criteria. A secondary purpose of
operating this pilot plant was to acquire a greater knowledge
of the C-G process. This part of the pilot testing was
limited to about 1 of the 18 months and was mostly incidental
to the primary purpose. The amount of dried sludge required
for test burning was originally estimated to be about four to
five tons of dry sludge. The actual sludge requirements
turned out to be on the order of 40 to 50 tons.
Operating costs for a full-scale C-G system are estimated to
be $73/dry ton based upon experience gained with the pilot
3-6
-------
TABLE 3-1
PROCESS PROBLEMS AND CORRECTIVE MEASURES
BURLINGTON INDUSTRIES C-G SYSTEM.
Process Problems
Action Taken
1. Lack of process and operational
knowledge including
instrumentation.
2. Operating parameters and standards
not established.
Initial lack of knowledge of the unknowns. 3,
For example, how dry the product needed
to be was unknown.
Excessive carrier oil loss -
9-12 gph vs 3 gph std.
Plugging of 4th and 5th stage
heat exchangers
Resolved with time and
exper ience.
Resolved from pilot data and
field experience - trial and
error.
Handled one by one. For examples, it
was important to maintain low water
content in the dry material as well
as a pH of about 5.0. High water
content would produce materials of
rock like consistency. Otherwise,
there was great difficulty in handling
the dry product through the centrifuge,
hydroextractor and conveyors. The
addition of biosludge to the lanolin
rich material between the fourth and
fifth stages of drying helped the
operations.
Tested all known sources for
losses. However, requires more testing
for oil losses. Testing other systems
for oil/water separation to improve oil
recovery.
Resolved with time and experience:
a. Determined proper oil to solids
ratio @ 7-9:1.
Additional heat exchangers.
Removed 90
exchanger.
ell in top of heat
-------
TABLE 3-1
PROCESS PROBLEMS AND CORRECTIVE MEASURES
BURLINGTON INDUSTRIES C-G SYSTEM (Cont'd).
Process Problems
Action Taken
Flow measurement - flow meters
malfunctioning due to type of
product.
to
i
00
Foaming caused by rate of throughput
in C-G System and detergent in
product. Any small amount of foam tends
to emulsify water with oil in the
oil/water separator operations
and reduce the capacity of separation.
To maintain adequate control of the
5th stage, the oil/solids ratio should
be from 8/1 to 10/1. It is very
difficult to sense the flow and
consistency of a solids to water to
oil mixture that continually changes.
Using ALPHASOMICS meter in sludge
line - not 100% satisfied.
Controllotron going into fault mode -
unable to contol system automatically
in the 5th stage. Okay where product
is not too dense. Tried density meter
without success.
The ratio is currently determined
manually each 1/2 hour by sampling and
separating into fractions with a bench
centrifuge. A suitable monitoring
device for automatic operation is being
sought.
Problem partially resolved by
requesting cut-back on use of
detergent when foaming occurs and
temporary one hour cut-back on flow
rate of wastes to be dried thorugh the
rate of wastes to be dried.
-------
TABLE 3-1
PROCESS PROBLEMS AMD CORRECTIVE MEASURES
BURLINGTON INDUSTRIES C-G SYSTEM (Cont'd),
Process Problems
Action Taken
8. Oil/water separator - capacity, and
carry over of oil and fines.
i
vo
9. Moisture/oil in hydroextractor
10. Centrifuge plugging.
11. Centrifuge erosion.
8. Present operation is drying 75 GPM
scouring waste (1/2 to 1% solids) and
4 GPM 20% solids bio-sludge from
finished wool cleaning. This is the
original design capacity, facilitated
by modifications made to give more
evaporation capacity. Both Burlington
and Dehydro-Tech are working to reach
higher capacities. Further increase in
throughput rate may be assisted by a
larger capacity for oil/water
separation and/or vapor foam
disengagement.
9. Problem overcome by improving moisture
content & quality of feed to
centrifuge.
10. Problem resloved by flushing with
carrier oil when shutting down.
11. Resolved problem by installing
removeable wear plates on inside of
casing.
Problems not resolved - 4, 6, 7, and 8
-------
TABLE 3-2
MECHANICAL AND PHYSICAL PROBLEMS AND CORRECTIVE MEASURES
BURLINGTON INDUSTRIES C-G SYSTEM.
Mechanical & Physical
Action Taken
to
I
1. Building size - too small.
2.* Material of construction -
CS vs SS piping, tanks, etc.
Leaks - erosion & corrosion.
3. Design changes - piping,
equipment, etc.
Inadequate capacity based on original
design for 80 GPM (37,000 Ib/hr) at
3.5% solids in the wool scouring waste
Actual solids content is 1/2 to 1%.
Needed capacity is for 90 to 100 GPM
(50,000 Ib/hr).
Hydroextractor -
a. Tines of rotor.
b. Tines drives.
c. Propane heater sink.
d. Pluggage in vapor lines.
e. Pluggage in scrubber bubble
trays.
f. Plate heat exchanger pluggage,
g. Scrubber bottom outlet
pluggage.
1. 62' x 52' - too late to correct.
2. a. Replacing as required.
b. Feeding ammonia to increase pH.
1.0#/hr. @ 5.0+ pH.
3. Altered/modified/removed as
required.
4. Installed additional small heat
exchangers to increase capacity.
a. Replaced tines with angle
iron.
b. Replaced "Carter" drives.
c. Made three pass - modified.
d. Increased velocity and in-
stalled manual rake between
deoiler and scrubber.
Modified piping.
e. Remove disk on tray opening.
f. Replaced with spiral heat
exchanger.
g. Modified bottom of scrubber
and installed C.O. sparger.
-------
TABLE 3-2
MECHANICAL AND PHYSICAL PROBLEMS AND CORRECTIVE MEASURES
BURLINGTON INDUSTRIES C-G SYSTEM (Cont'd).
Mechanical & Physical
Action Taken
6.
7
Agitator maintenance - bottom
bearing and shaft failure.
Pumps - erosion, corrosion.
Pump seals - mechanical by
Chesterton.
9. Pump impellers - wrong design •
should have been open impeller
10.* Leaks - piping.
11.* Acid system - leaks and pump
fa ilure.
12.* Vacuum system - replaced.
6. Replaced with larger agitator and
shaft - no bottom bearing.
7. Replacing and changing materials
of construction - Durco CD4M.
Fifth stage casing being replaced
(because of acid attack) with lower
cost CD4M steel compared with stain-
less steel. CD4M impellers holding up
very well.
8. a. Still experimenting.
b. Trying bellows vs spring
loaded seal - Chesterton.
c. Trying Sealol bellows type
seal.
d. Both seals holding up very well.
9. Replacing as required to prevent
wool, trash, etc. from plugging
impeller.
10. Replacing with SS. pH control with
ammonia feed.
11. a. Replaced piping with coated
piping.
b. Repair pump as required.
12. Replaced liquid ring pump 3 times-
last pump stainless steel.
-------
TABLE 3-2
MECHANICAL AND PHYSICAL PROBLEMS AND CORRECTIVE MEASURES
BURLINGTON INDUSTRIES C-G SYSTEM (cont'd).
Mechanical & Physical
Action Taken
13. Centrifuge.
14. Pelletizer
15. Solids handling
Moyno Pump (Sludge) - Excessive
pump maintenance and high
pressure drop.
13. Factory rebuilt in Jan. 1986.
14. Never tried to operate - solids
too dusty.
15. Replaced belt conveyors with screw
conveyors. Fair service now. Will
replace.
16. Replaced 4" piping with 6" piping.
General Notes;
*Items 1, 7, 10, 11 & 12 are corrosion problems that occurred in the condensate piping after
the 4th stage. These problems occurred due to sulfuric acid added at that stage to permit
separation of the lanolin. The subsequent addition of ammonia after the 5th stage should
eliminate many of the corrosion problems encountered in the past.
Ma intenance Costs;
Direct maintenance $25,000 to 30,000
1987 Overhaul type of maintenance, which
includes items such as vacuum pump
changes, piping, etc. This type of
maintenance should be reduced om
future years.
$30,000/year
75,OOP/year
$105,000/year
-------
TABLE 3-2
MECHANICAL AND PHYSICAL PROBLEMS AND CORRECTIVE MEASURES
BURLINGTON INDUSTRIES C-G SYSTEM (cont'd).
Mechanical & Physical
Action Taken
13. Centrifuge.
14. Pelletizer
15. Solids handling
Moyno Pump (Sludge) - Excessive
pump maintenance and high
pressure drop.
13. Factory rebuilt in Jan. 1986.
14. Never tried to operate - solids
too dusty.
15. Replaced belt conveyors with screw
conveyors. Fair service now. Will
replace.
16. Replaced 4" piping with 6" piping.
General Notes;
*Items 1, 7, 10, 11 & 12 are corrosion problems that occurred in the condensate piping after
the 4th stage. These problems occurred due to sulfuric acid added at that stage to permit
separation of the lanolin. The subsequent addition of ammonia after the 5th stage should
eliminate many of the corrosion problems encountered in the past.
Maintenance Costs:
1986 Direct maintenance $25,000 to 30,000
1987 Overhaul type of maintenance, which
includes items such as vacuum pump
changes, piping, etc. This type of
maintenance should be reduced om
future years.
$30,000/year
75,OOP/year
$105,000/year
-------
system. Including amortization of the $120 million capital
C°S^ 3f «??/*:, Percent ^terest for 20 years yields a total
cost of $248/dry ton. Currently, the LACSD pays $38/dry ton
for composting and $50/dry ton for landfilling.
Several operational problems were experienced in the pilot
plant because: 1) the pilot system was designed for
processing cattle feedlot manure and not sludge, and 2) the
age of the pilot system made maintaining operation quite
difficult. Under these circumstances, considerable
modification to the pilot plant was necessary. For example,
abrasion of pump seals was a significant problem as the
original seals lasted only four to seven days. This abrasive
problem was believed to be due to the alignment of the pumps,
the nature of the seal, and the abrasive nature of the
sludge. Tungsten carbide on tungsten carbide seals with an
oil flush between the seals was recommended as an interim
improvement for future systems but was not believed to be an
adequate long-term solution.
The LACSD was never were able to overcome poor centrate
quality in the recycle stream (i.e. more than expected
quantities of solids were carrying over into the centrate.)
A planned solution by LACSD is the use of a greater number of
smaller centrifuges operating at a lower speed (gravitational
force). Lower speeds also will reduce wear problems.
Others, including Dehydro-Tech, believe that higher
gravitational force is necessary to prepare a good cake and
to capture a sufficient amount of the fines.
Fires in the hydroextractor. also occurred. The
hydroextractor was not air-tight, and on two occasions, slow
smoldering fires occurred after the system was shut down.
Concerns regarding odors were also expressed. The pilot
system was 1/200 the size of the full scale plant to be
built, yet odor complaints were received as far as 1/4 mile
away. Since a full-scale C-G system is totally enclosed,
except for a limited number of exhaust vents, any possible
odor problems should be controllable.
By far the greatest concern expressed by the LACSD was with
the solids handling. For example, plugging o£__the_ hea t_
exchanger s_wa§_f e 11 to be a s igjiJULicant., prSSl£JiL~s.in£e__i t
occurred continually 3uHng~~their pilot operation. As
previously pointed out, Los Angeles City originally had
problems with plugging of the spiral heat exchangers with
wood chips. The city later felt that this problem was
overcome by screening larger solids pieces out of the test
materials being used and better process control to overcome
going into the "gummy phase". It is not yet known if other
problems will result in plugging of the spiral heat
'exchangers.
3-13
-------
system. Including amortization of the 5120 million capital
cost at eight percent interest for 20 years yields a total
cost of $248/dry ton. Currently, the LACSD pays $38/dry ton
for composting and 550/dry ton for landfilling.
Several operational problems were experienced in the pilot
plant because: 1) the pilot system was designed for
processing cattle feedlot manure and not sludge, and 2) the
age of the pilot system made maintaining operation quite
difficult. Under these circumstances, considerable
modification to the pilot plant was necessary. For example,
abrasion of pump seals was a significant problem as the
original seals lasted only four to seven days. This abrasive
problem was believed, to be due to the alignment of the pumps,
the nature of the seal, and the abrasive nature of the
sludge. Tungsten carbide on tungsten carbide seals with an
oil flush between the seals was recommended as an interim
improvement for future systems but was not believed to be an
adequate long-term solution.
The LACSD was never were able to overcome poor centrate
quality in the recycle stream (i.e. more than expected
quantities of solids were carrying over into the centrate.)
A planned solution by LACSD is the use of a greater number of
smaller centrifuges operating at a lower speed (gravitational
force). Lower speeds also will reduce wear problems.
Others, including Dehydro-Tech, believe that higher
gravitational force is necessary to prepare a good cake and
to capture a sufficient amount of the fines.
Fires in the hydroextractor also occurred. The
hydroextractor was not air-tight, and on two occasions, slow
smoldering fires occurred after the system was shut down.
Concerns regarding odors were also expressed. The pilot
system was 1/200 the size of the full scale plant to be
built, yet odor complaints were received as far as 1/4 mile
away. Since a full-scale C-G system is totally enclosed,
except for a limited number of exhaust vents, any possible
odor problems should be controllable.
By far the greatest concern expressed by the LACSD was with
the solids handling. For example, plugging of___th_e .heat_
exchangers -was felt to be_a^3JLULLfl a n t pAfl&Iam^ 1 nc e_.i t
occurred ^i^l^^^^^J^ Anally had
previously pointed out, LOS nuycj-c , -J • ,_,_
^ ,. ^..'V, „!,„.„< no of the spiral heat exchangers with
problems with P^1" later felt that this problem was
wood chips. The city la* « . of the fcest
bv screening .target auj.±<^ c
Lused and b.tt.r prcc... =»tr.l^o ovrcoj.
'
exchangers.
3-13
-------
The LACSD also had plugging of one-inch lines in the pilot
operation, yet the spiral heat exchanger specifications for
all four systems call for openings of 5/16 to 7/16-inch.
Backflushing capability for the heat, exchangers could help
alleviate this problem as could prescreening with a
1/10-inch mesh screen and/or use of a Muffin Monster (TM)
with a 1/8-inch mesh screen. There was also discussion about
the possible need for a Reitz mill (a fine grinding impact
type mill) , but the possible need for such a device was not
resolved. A Reitz mill would add about 150 HP to the energy
requirements of the system for the City of Los Angeles.
LACSD also recommended that, if possible, a municipality
consider using the heavy oil system and eliminate the
hydroextractors. The LACSD did not believe that there was
sufficient knowledge about how light oil systems will perform
and that new systems should not be designed until further
operational experience is gained from the four plants already
in start-up or construction. Foster Wheeler engineers later
noted that due to the time required for design, the
experience gained at the four existing municipal systems/
could be incorporated into the design of any new systems|
before design of these new systems was completed.
3-14
-------
SECTION 4
RECOMMENDATIONS
4.1 Introduction
Numerous means to improve the design, construction, and
operation of the C-C system were discussed during the
seminar. A brief discussion of the major recommendations is
presented herein.
4.2 Responsibilities for Design and Construction
The CG systems that appear to have had the fewest problems
during design and construction are those systems which have
had the fewest groups involved with these tasks. The LACSD
and Ocean County systems, which have either had one firm or
agency in control of the C-G portion of the project or have
utilized special management techniques, have had fewer design
and construction problems, to date, than the City of Los
Angeles and Mercer Co. Th.e City of Los Angeles has had by
• far the most complex arrangements (Tables 4-1 and 4-2) 'with
responsibility greatly divided between consultants and
contractors durj.nq the__ various phases of design and
construction.
These experiences emphasize the essential need for a single
party to have overall control of th_g, *<=•* ig" — a-R-d construction
to ensure that meaningful Interface between all parties
involved with a C-G system occurs. Appointing one design
firm as the prime A/E firm for a sludge treatment plant could
minimize potential problems and help ensure that good design,
construction, and start-up procedures are used.
Foster Wheeler is the only engineering firm authorized by
Dehydro-Tech for the design C-G systems of with larger than
30 dry tons per day drying capacity. Typically, however,
other engineering firms are responsible for the design of
other components of a total sludge treatment plant which must
mesh closely with the C-G system. Vertical division of
design tasks appear to be superior to horizontal division.
With vertical division one firm designs ^portion of the
plant from the foundations to the roof. JU^e-C^ty-^Las.
espo nubilities?) ajid
"
with vertical division "of
between d^er^t-d^ae^
and construction contractors must be carefully managed and
controlled.
4-1
-------
TABLE 4-1
FW SCOPE OF WORK
CARVER GREENFIELD FACILITY
STEP II DESIGN
A-E Consultants
INTERFACES
-Wet Cake
-Dry Product
-Sewage oil
-Utilities
-Electrical
PROGRESS DESIGN
SITE DESIGN
CIVIL DESIGN
STRUCTURAL DESIGN
| ARCHITECTURAL DESIGN
N)
MECHANICAL DESIGN
ELECTRICAL DESIGN
INSTRUMENT DESIGN
CENTRAL CONTROL SYSTEM
SCALE MODEL
BEGIN DESIGN
1001 DESIGN SUBMITTAL
CONSTRUCTION AWARD
OVERALL PROJECT COST
IN MILLION $
C-G FACILITY COST
IN MILLION $
CITY OF L. A.
Montganery-Parsons
(M-P)
CONVEYOR DISCHARGE
SOLIDS COOLER OUTLET
PIPE CONNECTION @ B.L.
OSBL
480V HOC
FW / DTC
M-P
M-P
M-P
LAC
M-P / FW
M-P / FW
M-P / FW / CONTRACTOR
M-P (MP)
AFTER DESIGN (PR MODEL)
FEBRUARY 1982
JANUARY 19B3
OCTOBER 1983
Bid C$15ff)
To date $216
Bid \$2o)
LACSD
None
CONVEYOR DISCHARGE
STORAGE SILOS
STORAGE TANKS
SELF-CONTAINED
12KV FEEDER
FW / DTC
FW / LAC
FW
FW
FW
FW
FW
FW
LACSD (MP)
DURING DESIGN
(COMPLETE MODEL)
OCTOBER 1983
JUNE 1985
MARCH 1986
$120
$50
MERCER COUNTY
Clinton Bogert Assoc
(CBA)
CONVEYOR DISCHARGE
PELLETIZER DISCHARGE
BOILER/TURBINE GENERATOR
SELF CONTAINED
13.2KV SWITCHGEAR
FW /DCT
CBA
CBA
CBA
C fl A
FW
C B A / FW
FW / CONTRACTOR
FW (A)
NONE
MARCH 1982
SEPTEMBER 1983
AUGUST 1984
$35
$12
OCEAN COUNTY
Havens t, Emerson/
Brown & Ca Id well
(H&E / B&C)
PIPE CONNECTION
SOLIDS COOLER OUTLET
PIPE CONNECTION
OSBL
480V MCC
FW / DTC
H&E / B&C
FW
FW
H&E / B&C
FW
H&E / B&C / FW
FW
H&E / B&C (A)
NONE
JUNE 1984
JUNE 1985
BIDS RECEIVED
FEBRUARY 1986
$27
$6
Wheelec
DTC - Dchydro-Tcch Corporation
-------
TABLE 4-2
FW SCOPE OF WORK
CARVER GREENFIELD FACILITY
STEP III SERVICES
CITY OF L. A.
A-E CONSULTANTS M - P
CONSTRUCTION MANAGEMENT M - P
TECHNICAL REVIEWS FW
FW FIELD REPRESENTATION PARTIAL
DURING CONSTRUCTION
PROCESS OPERATING MANUAL FW
0&M MANUAL M - P
(MECH CATALOGS)
OPERATOR TRAINING M - P w/FW ASSISTANCE
COMMISSIONING M - P w/FW ASSISTANCE
INITIAL START-UP M - P w/FW ASSISTANCE
PERFORMANCE TESTING FW/ DTC
PRIME ENGINEERING M - P
CONSTRUCTION AWARD OCTOBER 1983
MECHANICAL CONSTRUCTION SEPTEMBER 1985
COMPLETION (SCHEDULED)
MECHANICAL CONSTRUCTION JUNE 1987
COMPLETION (FORECAST)
LACSD
NONE
LACSD
FW
FULL TIME
FW
CONTRACTOR
FW
FW
FW
FW
FW
MARCH 1986
SEPTEMBER 1988
SEPTEMBER 1988
TRENTON
C B A
C B A
FW
PARTIAL
FW
C B A / CONTRACTOR
DTC
DTC
DTC
DTC
C B A
AUGUST 1984
JANUARY 1987
JULY 1987
OCEAN COUNTY
HS.E / B4C
0 C U A
FW
NONE
FW
IJ&E / BiC
H&E / B&C
W/FW ASSISTANCE
FW / DTC
FW / DTC
FW / DTC
H&E / B&C
SPRING 1987
FALL 1989
FALL 1989
1. CBA: Clinton Bogert Assoc.
2. DTC: Dehydro-Tech Corporation
3. FW: Foster Wheeler
4. H&E / B&C: Havens & Bimerson/Brown & Caldwell
5. M-P: Montgomery-Parsons
-------
All design firms involved must understand the process
K^Vi"' *he Pr°c?ss strong points and limitations, and
the influent and effluent requirements. The design firm
responsible for conventional technologies must co-ordinate
with the C-G design firm to ensure that the sludge treatment
and initial dewatering processes are compatible. For
example, multiple vapor recompression C-G systems are thought
to be more economical for sludges with a high water content
(e.g., 5 to 10 percent solids); whereas, multiple effect
systems are thought to be more economical for dryer sludges
(e.g., 15 to 20 percent). The designers of the C-G system
must, therefore, know the projected water content of the
sludge to be produced by the conventional treatment system.
Conversely, the designers of the C-G system must provide
other design firms with information on any C-G process
streams which will be recycled to the treatment plant. Thus,
responsibility for design should be clearly defined.
4.3 Design_
4.3.1 Philosophy
The HERS was originally designed to provide maximum energy
recovery and operational flexibility. As a result, the
system components were highly integrated. Operational
problems in one unit could significantly affect the operation
of all other units. The complexity of operation was thereby
increased. The HERS C-G system is currently being
selectively redesigned to be less integrated. Because__the
'C-G process is new for wastewater .treatment, the importance
of operational simplicity versus maximum energy usage and
recovery should be considered..
Safety should also be a primary concern in design. Explosion
venting and nitrogen blanketing were added to the HERS system
after construction had begun. HERS engineers also added
instrumentation and redundancy in both equipment and
protective safeguards. The LACSD system has been designed to
shut down in a fail safe mode should control problems
develop. Unplanned combustion of solids in the system has
occurred in both pilot and full-scale systems; thus, safety
is important.
Scheduled downtime for maintenance should be considered. The
Burlington Industries and two New Jersey systems do not or
will not operate continuously because of feed availability.
These systems do not operate at least two days per week
(generally weekends). Regular maintenance can thus be
conducted without disrupting solids handling at these
facilities. Time has not been allotted for routine
maintenance at either of the Los Angeles systems. The lack
4-4
-------
hoVH scheduled maintenance could become a problem
should operational problems arise. To alleviate this
potential problem, redundancy has been provided so that, at
least in theory, portions of a given process train can be
shut down for maintenance while continuing to operate the
remaining portions of the train. Providing for at least 2
days sludge storage prior to the C-G process should also be
considered.
Any municipal wastewater treatment authority considering the
C-C technology, should review the experience performance of
the four municipal C-G systems discussed in this report.
Visits to and detailed discussions with as many of these four
system owners and engineers as possible should be conducted
so that the experience gained from these systems can then be
incorporated into their systems. Accurate operation and
maintenance cost data should also be available for use in
determining real plant costs. All these steps are essential
in weighing a possible committment to using the C-G
technology.
4.3.2. Plans and Specifications
Preparing detailed plans and specifications is critical to a
C-C project. Time constraints limited the time for
specification preparation and TPVJPW of i-hg LA HERS project.
Pages were inadvertently left out of the project
Specifications whirh resulted In piping prnblpma.
Improvements in the wording of the specifications could also
have been made. For example, the specifications for the
spiral heat exchangers were intended to have been written
such that a 5/16-inch plate, spacing would be the minimum
opening, including a tolerance of 1/8-inch. Instead,
5/16-inch became the nominal size for the opening, plus or
minus 1/8-inch. The minimum opening of the heat exchanger
could have been 3/16-inch, while the opening on the sludge
grinder is 1/4-inch. This could cause plugging of the heat
exchanger. While the openings in spiral heat exchangers
could have been made larger (over one inch in diameter) to
avoid possible plugging, the larger openings would result in
larger less efficient units for exchanging heat.
Fortunately; ^° mannfartnrer of the spiral heat exchanger.
made the City OJL r.ns Anaeles exchangers with a 5/16 to
7/16-inch opening.
4.3.3 Equipment Selection
With new applications of a technology, reliable process
nn •; pmant- i«T critical since experience with operations is not
available—to help compensate for equipment limitations.
Desicmers of the system recommended that,if possible, sole
source purchasing should be conducted for some process
4-5
-------
equipment. The construction contractor should also be made
aware that a particular manufacturer of a piece of equipment
is preferred if the design of the system (e.g., piping) is
based upon this equipment. Otherwise, the equipment
purchased (whether or not it is the least expensive to
purchase) may not fit into the space allotted and could
require redesign of associated portions of the system.
Delivery of process equipment should be scheduled carefully.
Othewise, process equipment may arrive months or years before
it can be used. If problems develop upon operation,
difficulty may be encountered in having the manufacturer
honor the warranty.
Early operational experiences have shown that problems may
occur with pump seals; flow meters; level, viscosity and
other sensors; valves; oil/water separators; centrifuges; the
hydroextractor (de-oiler) and heat exchangers. Solids
abrasion of the seals causes relatively rapid breakdown.
Pump manufacturers may not provide any assistance should seal
problems develop. Instead, the manufacturers may refer the
owner to a seal manufacturer. H_iqher gnaii.-t-y pumps wi i-h
hardened surfaces are apparently no^H^r?. Rvpgr -Imgnhai- i on
w^th pump seals is si-itJ — underway at Burlington— aud — ttue_Ci±-y_
of Los Angeles. Future systems should contact the existing
s^ys terns t"o determine which types of seals full-scale
operational experiences have shown to be effective. Double
mechanical (tungsten carbide on tungsten carbide) seals
appear to offer the most promise at this time.
Accurately measuring solids flow in the system is very
important and difficult; thus, mass flow meters may be needed
for automated control along with manual measurement of the
solids to oil to water ratios. Doppler-ef feet flow meters
may also prove to be effective.
Level control _jn j-ho ^ygporn*""1"1 — i-s — v_g>ry important for
obtaining good heat transfer. Initially, level sensors in
the HERS effects provided inaccurate level data. As a
result, process efficiency decreased. Replacement of the
fluid in the level controllers corrected this problem at the
HERS.
An oil-water separator, hooked to an evaporating system, is
needed for effective operation that will routinely permit
essentially complete recovery of the carrier oil (not more
than about 100 ppm residual free oil in the water). The
t-n ^ate have been (a) to avoid solids Crtrrv-nvftr to
at-gr separate*- ^rom thfa evaporator,, and — (h) reliable
sepIrTtor. This solids carry-over at the
Cit7~of Los Angeles has prevented the establishment of a good
interface in the separator because of the formation of an
4-6
-------
emulsion. Separation has also been inconsistent (e.g., at
Burlington) where solids carry-over to the separator has not
been a problem.
If oil is not properly separated and is carried over into the
water, the oil goes back to the head of the wastewater
treatment plant and is thus wasted. If water is not
separated from the oil, the water would go with the carrier
oil to the fluidizing tank and could cause pumpability
problems and plugging of the evaporators and heat exchangers
due to the formation of a gummy phase.
Both high- and low-g centrifuges have been recommended to
reduce carryover of fine solids into the centrate. It is not
really know at this time which will work best. Remove able
wear plates shoulfl b^ instil grf in t-he. ^f-ntri f 'NflS b.^aiiRP- "f
the abrasive nature of <-ho maj-oripis ben ng._ pumped . Sole
source purchasing of the heat exchangers as well as
centrifuges is recommended by the design engineers because
the both pieces of equipment are very critical to successful
operation.
Normal pressures in the spiral heat exchangers initially
caused deflection of the endplates at HERS, resulting in
mixing of solids in the two flow paths. Manufacturer
supplied replacement gasketing and strenghtened endplates
have corrected this problem. Pressure testing of the heat
exchangers is recommended to ensure that leakage between flow
paths does not occur. Only one flow path should be
pressurized during testing.
Redundancy should be nnnsiripred _ fox — any components — £r_Q_n£ — Lo
failure or for which failure could pose a safety risk.
Specifications for equipment should HP rar^fnlly written for
fyTture systems. Qpgrati,9nal experiences acquired wii*"1"1 — tke-
existing systems will provide the data needed to prepare
better specifications.
4.3.4 Plant Model
Constructing a scale model of the C-G system is highly
beneficial. Not only can the model be used for operator
training, but also for checking the design and to assist the
contractor in construction. Burlington Industries found that
the $28,000 required to construct the model saved at least
$50,000 in construction. Models are very useful for ensuring
(the process equipment and piping will fit in the available
'space. Problems with the equipment not fitting in the
available space have occurred at systems where a model was
not built.
4-7
-------
4.4 Contractor Selection and Construction Activities
As with any construction project, having a qualified
contractor is very important to the success of a project.
Many contracts were required to construct the Los Angeles
HERS project within the time constraints imposed by the
Consent Decree. One of the prime contractors, although not
the C-G contractor, performed very poorly. Because the KERS
was highly integrated, the failure of this one contractor
impacted all components of the HERS project, including the
C-G system. Under such conditions, total integration of the
system was actually a detriment. This contractor is no
longer involved in the project. Also/ no liquidated damages
were placed on mp°<"irLg—the . project—g/*hA . f im«a rpnsfra i nt-g
required for HERS. Tp r^t-rngpor-> .r—tk«—actual—elapsed—t-ime
f-r-om process selection j-n sl-arft-^p ._couJLd. perhaps have be&n
reduced had a more realistic schedule been acceptable fr.om
the beginning.^ A more realistic schedule would have
permitted a more careJiuJLLy- controlled design and construction
approach.~~'~
The Los Angeles CSD, on the other hand, was not under any
legal time contraints and was able to proceed at a slower
B.f3CR> Fewer prime construction contractors are involved in
the construction of the LACSD system. . Mercer County
expressed concern that their contractor did not fully
comprehend the project. As a result, the plant was only 85
percent completed by the original' project completion date.
It was felt that hiring a contractor familiar with chemical
plant type construction may have eliminated many of Mercer
County's construction' delays.
Ocean County has attempted to avoid contractor and start-up
problems by writing the original bid documents in such a way
that Dehydro-Tech, the construction contractor, the design
engineers, and the County are sharing responsibility for the
project until all testing of the completed system has been
conducted and approved. Payments to the contractor for
start-up work are to be made based upon actual services being
performed, rather than upon a lump sum for the entire task.
Ocean County believes it saved an estimated $500,000 in the
bids it received by using this approach.
A general comment made during the workshop was that when
compared to a petrochemical plant, construction of a C-G
system is relatively simple. Therefore, other treatment
authorities which utilize the C-G process should consider
selecting contractors with petrochemical plant construction
experience.
4-8
-------
KERS personnel stressed that key individuals in the proiect I
must have time to review the plans and specifications in
detail. The LACSD system required approximately %So
drawings; thus, a significant time for review is required for
a C-G system. These same individuals should be present or
available during construction to answer contractor's
questions and routinely monitor the contractor's performance.
4.5 Operation
4.5.1 Personnel Requirements
The number of individuals required to run a municipal
wastewater C-G system is not yet accurately known. One
operator was reported to be sufficient for operating three
C-G trains in the rendering industry. Without computer
control the Burlington Industries' system requires two people
per shift, with an additional person on day shift to assist
in routine maintenance and solids truck loading. The LACSD
system will be entirely computer controlled because the
operation of the process is felt by the LACSD to be too
complex for manual operation.
Differences in opinions on the educational and/or experience
level needed by an operator to run a C-G system were
expressed. The opinion most commonly expressed was that
refinery or petrochemical operators and supervisors would be
well qualified to operate a C-G system. A significant
difference in risk exists between a hot hydrocarbon leak from
a C-G system versus the relatively innocuous spills of
wastewater and sludge with which most wastewater treatment
plant operators are familiar. Regardless of the personnel
background, it is important that all personnel receive
adequate training and that they are able to comprehend the
care that needs to be exercised in working with systems that
involve the use of hot hydrocarbons.
4.5.2 Operator Training
A key point made during the discussions was that the C-G
process is, in effect, a petrochemical type plant and not a
conventional wastewater type treatment process. Safe and
efficient operation of the C-G process requires qualified
operators. Adequate training must be provided. Operators at
the Hyperion plant were scheduled for a total of eight days
of training from Foster Wheeler. There was additional
training on specific equipment by vendor representatives. In
retrospect, some individuals associated with the Hyperion
system believed that more specialized training time would
have been beneficial. Various opinions on the amount of
training required were voiced and ranged from three weeks to
4-9
-------
three months. Allowing operators to observe the operation of
other systems would be highly beneficial.
To assist in training, a physical model of the C-G process
should be constructed. Due to the relative complexity of the
system, a municipality should also consider contracting or
negotiating for (as part of the design cost) full-time,
on-site technical support for the process during start up.
Finally f a pilot! plnni; gyghg^ ghpuld be operated at least on
£hose new untried components of a_£^G___sxjs££ESL_to ,Determine
they will interact with, t-hg entire pronpfici. to better a'ssu'Fg
their suitability and to better acquaint plant personnel with
the^ design of the process. Treatment plant personnel will
then be able to have better input into the system design as
well as have acquired valuable training in operating the C-G
system.
4.5.3 Start-up and Solids Recycle
When a plant first begins operation, sufficient solids to
avoid a gummy phase will not be present. Thus, another
source of solids will be needed for start-up. Composted
sludge is one material recommended for use during start-up if
the compost does not contain particles larger than the
clearances in the process equipment. Fine grinding and./or
Screening of the ^ompn ^ figr] gj^^qo may a 1 g n h f*—noo^o(-j 1-n
prevent wood__pr fjh^rs _fmm plugging heat exchangers. Other
suitable materials might include heat-dried sludge products
such as milorganite or dried solids from other C-G systems.
Changes in particle size occur as the solids pass through the
C-G system. Particle size may grow through the first three
evaporators. In the fourth evaporator, particle size may
decrease if forced circulation evaporation is used. Under
routine operating conditions particle size reduction is not
anticipated to be a significant problem. During start-up
however, solids will recycle through the system several times
until the solids inventory increases. A significant amount
of fine solids can thus be generated. Some of these fines
will also be removed from the system with the centrate from
the centrifuges and from the de-oiler into the vapor lines.
The fines can also be drawn into the vapor vent lines from
the hydroextractor.. Then, if cooling of the pipes occurs,
these fines stick to the interior pipe walls, and plugging of
the vapor vent lines or fires could result. _ The city of LA
is solving the problems by re-piping to gain an increased
vapor flow velocity (to a minimum of 50 feet per second)
through a shorter more direct path They are also installing
a dust trap and nitrogen gas blanketing as an extra
precaution against solids build up and auto-oxidation.
4-10
-------
4.5.4 Cen trate Qua1i ty
K:hen the LA HERS was initially designed, Foster Wheeler
contacted the centrifuge manufacturers to discuss reasonable
design specifications for their equipment. The manufacturers
stated at that time that a c^nt-rate solids concentration n€
0.5 percent solids was feasible. DurJ_nji_j3ldd_LcLg- of t-h^
project, however, manufacturers would only guarantee 1
percent solids. As_ a result/ the centrate qualH-y was 1 PSH
than originally designed. One percent solids concentration
are now expected under routine operating conditions; however,
solids concentrations in the centrate may be even higher
during start-up due to increased fin^s r^nH-ir^—from
excessive recycling.
Failure to capture sufficient solids (say only 97 percent
compared with 99 percent) means that there will be three
percent instead of ong pprr-pnt fines in the sewage and
carrier oil going to i-he» ft a_sJa_« t-ii i . AS the carrier oil is
flashed off to the sewage oil so that it can be recycled back
into the evaporating process, the solids in the sewage oil
become more concentrated. The resultant solids content would
then be about 50 percent instead of 20 percent. Too high a
solids content is not pumpable and the system plugs up and
increased loss of carrier oil could occur. .Future systems
should pay particular attention to t_he performance ... nf I-HP
centrifuges and_the effect t;o _ centrate^ ,q^,a,li ty QQ frihpi...iiPY'' «-T-ng
systems.
4.5.5 Excessive Loss of Carrier Oil
Carrier oil losses are thought to occur primarily (a) into
wastewater during oil-water separation, (b) into the sewage
oil from the flash stilling process, and (c) into the solids
product from the de-oiling process. Excessive loss of the
carrier oil can be expensive. Improvement in the equipment
and operation to recover more of the carrier oil is
necessary.
4.5.6 Use of the Dried Sludge
It is important to anticipate possible, limitations on
ultimate use and disposal of the dry end product, including
ash. in August 1987, the EPA was scheduled to issue new
draft comprehensive sludge management regulations. The
constituents and concentrations of these contituents present
in the sludge will determine the available options for sludge
disposal.
Dehydro-Tech engineers stated that oil soluble organics
should be removed from the sludge in the C-G process.
Orqanics including sewage oil would be extracted from the
sludge by the carrier oil. The removed sewage oil can be
-------
monih „ I concentration in the HERS is routinely
monitored to ensure that the PCB levels are acceptable for
incineration xn their system. Pathogens are killed by the
high temperatures and dehydration because of the lonq
retention time of the sludge in the C-G system.
Metals concentrations may be of concern as non-volatile
metals should remain with the solid particles. Volatile
metals, such as lead and particularly mercury, may be removed
from the sludge, and air pollution control devices capable of
removing these metals may be needed depending upon local air
quality standards and the metal concentrations in the sludge.
The dried sludge from the HERS is currently classified as a
hazardous waste by California's Wet Extraction Test (WET)
Procedure due to cadmium; although, the levels of cadmium
detected were not hazardous by the U. S. EPA Extraction
Procedure (EP Toxicity) Test.
Processing differences can have appreciable influence on the
"degree of hazardousness" of the C-G end product. The
quenching of the ash with water, which will occur at HERS,
should reduce readily leachable metal levels per unit volume.
The ash from LACSD is not expected to fail the WET Test
because the lime used in the process will decrease metal
leachability. Future plants will have the option to produce
either a non-leachable slag or an ash.
4.5.7 Process Monitoring and Control
Parameters required for process monitoring are reported to be
few. pjarc-pnt- so|,irlg i .q f ppn^; ^d fro be the only routine
monitoring parameter for the influejit raw... jE,sed. Effluent
from the fluidizingtan'K should be monitored for the
solids/water and oil/solids ratio. Addback should be
monitored (to assure the proper water/solids ratio) to avoid
the gummy phase. Temperature and pressure can be used to
monitor the evaporators, while percent oil should be
monitored in the dried product.
Actual process control is more complicated. Numerous devices
such as flow meters, viscometers, level controllers, and
pressure and temperature gauges are all needed to better
control the operation of a C-G system. Problems with all of
these measuring devices or sensors have been experienced.
For example, there is difficulty in being able to
automatically measure and control the exact amounts of
solids, water and carrier oil being pumped and mixed at
critical locations in the C-G system. This is particularly
critical because excessive carrier oil in the system will
reduce the capacity because the excess oil must be
volatilized and recycled. Too little carrier oil may likely
4-12
-------
result in attaining a water content in the solids and oil of
about 30 percent which might not be pumpable due to formation
of "gummy phase." Manual half-hourly sampling and
determination of the solids-to-oil ratio is being
successfully used at the Burlington Industries C-G facility
to aid in process control.
Many of the minor problems, such as inaccurate readings from
the level controllers, have been corrected (e.g., by
replacing the fluid in the level controller) . With time and
additional operating gypprion^o mrjny existing problems will
be_ corrected,, uh n o *t—^k^—a^m
-------
that project and the already invested federal funds. Whether
or not those changes are federally funded is entirely a
separate issue and immaterial with respect to the grantee's
obligation to make the changes at the time that they become
apparent and necessary.
A grantee that fails to make necessary changes when they
are apparent, would be liable for charges of mismanagement or
negligence in the handling of federal funds. Such a charge
would seriously impair further federal funding and question
those funds already invested in the project.
M/R funding
which:
is only potentially available for a project
has completed construction,
has been accepted by the grantee,
has begun a performance period, and
has failed to meet performance standards.
Until these conditions have been met, there is no
demonstration that the I/A technology has failed and that M/R
funding should be made available. There also are additional
State requirments which place limitations on M/R funding
availablity even if these conditions are met.
4.7 Future Meetings
The participants in the seminar unanimously agreed that the
seminar was very worthwhile. Many participants recommend
that a follow-up seminar be held after more operational
experiences is acquired. Because of the problems still
unresolved, many wanted to have another meeting after 6
months, or approximately in September 1987. However, March
1988, was ultimately proposed as a more suitable time for a
follow-up meeting.
4-14
-------
SECTION 5
REFERENCES
1. Hyde, H. c., 1984. Technology Assessment of
Carver-Greenfield Municipal Sludge Drying Process/
EPA-600/2-84-200 (NTIS PBS5-138634) , Water Engineering
Research Laboratory/ Cincinnati, OH.
2. Crumm, J. c. and K. A. Pluenneke, 1984. "Development of
an Efficient Biomass Drying Process and its Commercial
Use for Energy Recovery." Presented at the Institute of
Gas Technology Symposium on Energy from Biomass and
Wastes, Orlando, Florida, February 1, 1984.
3. Walters, S., 1985. "Benefits from Biowaste." Mechanical
Engineering, pp.70-75, April 1985.
5-1
------- |