EPA-450/2-78-013
April 1978
COST AND ENGINEERING STUDY -
CONTROL OF VOLATILE
ORGANIC EMISSIONS
FROM WHISKEY WAREHOUSING
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
-------
EPA-450/2-78-013
COST AND ENGINEERING STUDY
CONTROL OF VOLATILE
ORGANIC EMISSIONS
FROM WHISKEY WAREHOUSING
Emission Standards and Engineering Division
Chemical and Petroleum Branch
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and ^ aste Management
Office of Air Quality Planning and Standards
Research Triangle Park. North Carolina 27711
April 1978
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - in limited quantities - from the
Library Services Office (MD~35) , U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina 27711.
Publication No. EPA-450/2-78-013
-------
TABLE OF CONTENTS
Page
1.0 Introduction , 1-1
1.1 Emission Source Description 1-1
1.2 Control Device Description 1-2
2.0 Whiskey Warehousing and Aging 2-1
2.1 Barreling and Warehousing - 2-2
2.2 Mechanisms of Aging 2-3
2.3 Warehouse Operation 2-8
2.4 References 2-16
3.0 Volatile Organic Emissions from Whiskey Warehousing ... 3 .... 3-1
3.1 Emission Source Description , 3-1
3.2 Whiskey Warehousing Emission Factors , . . . . 3-2
3.2.1 Emission Factors from IRS Data „ 3-2
3.2.,2 Emission Factors from Individual
Distiller Data . . . , 3-4
3.3 Emission Inventory a „ . . , . 3-12
3.4 References 3-14
4.0 Warehouse Emission Control , 4-1
4.1 Carbon Adsorption - System Description = . . . 4-1
4.2 Carbon Adsorption - Cost Analysis , 4-2
4.3 Carbon Adsorption - Feasibility , . ., . 4-6
4.3.1 Effects on Whiskey Quality 4-8
4.3.2 Re-use of Recovered Alcohol 4-10
4.3.3 OSHA Standards, Energy, etc , = .... 4-11
4.4 Carbon Adsorption - Warehouse Tests - 4-14
4.5 Alternate System of Aging 4-20
4.6 Control of Barrel Soakage Losses 4-22
4.7 References 4-26
-------
LIST OF TABLES Oano
r _ojj_S
Table 1-1 Control System Costs 1-3
Table 2-1 Statistical Data of Whiskey Maturation Study
by Liebmann and Scherl o 2-5
Table 2-2 Characteristics of American Whiskies at Various
Ages 2-5
Table -2-3 Warehousing Operations 2-15
Table 3-1 Losses, Withdrawals and Stocks of Whiskey
for the U.S 3-3
Table 3-2 Barrel Soakage Losses 3-3
Table 3-3 Evaporative Losses During Storage 3-6
Table 3-4 Computed Annualized, Cumulative and Incremental
Losses 3-6
Table 3-5 Warehouse Barrel Age Distribution 3-9
Table 3-6 Summary of Emission Factors for Whiskey Warehousing 3-11
Table 3-7 Total Emission Estimate by State 3-13
Table 4-1 Recovery System Costs » > 4-7
Table 4-2 Distilled Liquor Sales and Industrial Alcohol Use 4-12
Table 4-3 Carbon Adsorption System Data, 1960-1968 4~15
Table 4-4 Cost Calculations 4-17
Table 4-5 Control System for Barrel Soakage Losses - Warehousing 4-25
LIST OF FIGURES
Figure 2-1 Effect of Maturation on the Physical and
Chemical Characteristics of Whiskey 2-6
Figure 2-2 Mechanisms of Whiskey Aging 2-10
Figure 2-3 Diffusion through Barrel Staves in Whiskey
Aging 2-11
Figure 3-1 Emission Rate Relationships in the Whiskey Aging
Process • 3-7
-------
UNITS AND CONVERSIONS
Listed below are abbreviations and conversion factors for the metric
units in this report and definitions for non-standard units associated with
whiskey production.
Metric Unit (AbbreviationJ Equivalent
1 meter (m) = 39.37 inches
=3.28 feet
_r>
1 centimeter (cm) =10 meter
= 2.54 inches
1 hectare (ha) = 105 m2
= 2.47 acres
1 kilogram (kg) = 2.2 pounds
1 metric ton (MT) = 1000 kilograms
= 2200 pounds
Unit Definition
proof gallon (pg) one U.S. gallon of 231 cubic
inches containing 50 percent by
volume ethanol or any volume of
liquid containing an equivalent amount
of ethanol. A proof gallon thus
contains 1.5 kilogram of ethanol.
proof twice the volume percent ethanol
in a liquid. The number of proof
gallons in a gallon of liquid is the
proof divided by 100.
-------
-------
1.0 INTRODUCTION
The Environmental Protection Agency is currently providing technical
assistance to the States and local jurisdictions on industries that emit
significant quantities of air pollutants in those areas of the country where
National Ambient Air Quality Standards are not being attained. This document
is related to one such industry, whiskey warehousing. It is a significant source
volatile organic chemicals (VOC) in the area where the industry is concentrated,
Kentucky, Illinois, Indiana, and Tennessee.
1.1 EMISSION SOURCE DESCRIPTION
In producing whiskey, alcohol distilled from fermented grain is stored
in charred oak barrels for periods of four to eight years or more. During
this period, the alcohol absorbs, and reacts with, constituents in the
barrel wood and gains the distinctive taste and aroma of whiskey. This process
is known as aging or maturation. During the aging period, ethanol and water seep
through the barrel and evaporate into the air. Also when the barrels are emptied
to bottle the whiskey, ethanol and water remaining in the barrel wood evaporate
into the air. These last two phenomena are the major sources of VOC emissions in
whiskey production.
Based on changes in the proof and liquid volume of whiskey during aging,
,^>i>Q^
an emission factor of 3.2 kg/barrel-yr. was computed. On the basis of production,
the emission factor is .2kg ethanol/kg produced. Based on an estimated 10,260,000
1-1
-------
barrels stored in Kentucky, Illinois, Indiana, and Tennessee, the total yearly
emission of VOC from whiskey warehousing is 32,800 MT/yr for the four State
areas.
1.2 CONTROL DEVICE DESCRIPTION
The method investigated for control of emissions both during aging and
from barrel soakage after aging was carbon adsorption. Control of emissions
during aging would involve closing the warehouse and ducting exhaust from the
facility through a carbon adsorption unit. Control of barrel soakage losses would
involve placing the empty barrels in a closed warehouse ducted to a carbon adsorption
unit. These control methods are estimated to reduce emissions by 85 percent.
The efficiency is limited by the need to design and operate the system in a
manner that will not affect whiskey quality and by the physical difficulties in
drying the saturated barrels.
The applicability of these control systems is determined by two factors:
1. the cost of systems and
2. the system's effect on whiskey quality.
The cost of the system for controlling losses during aging for three of the
six cases studied is shown in Table 1-1. Also shown is the cost of controlling
soakage losses by storing the empty barrels in a warehouse. As seen in the table,
an important factor in the systems' cost is the credit for the recovered
alcohol. The recovered alcohol can be redistilled to a product for which
sufficient markets exist to use the amounts recovered; however, very few distillers
have the equipment required for this redistillation. Thus, distillers would have
to transport the recovered alcohol in crude form or install the necessary distillation
equipment, options which significantly reduce the credit shown for the recovered
alcohol.
1-2
-------
Table 1-1
CONTROL SYSTEM COSTS
Aging Loss Control Soakage Loss Control
Warehouse Size, Barrels 20,000 50,000 100,000 50,000
Annual Capital Costs $9,960 $15,410 $31,700 $71,000
Annual Operating Costs $11,980 $17,280 $26,010 $58,710
Annual Credit, $13,610 $54,440 $68,050 $55,150
Recovered Alcohol
Net Cost (Return)/yr $8,330 $(21,750) $(8,340) $74,560
Cost/Final Proof Gallon 3,0
-------
occurring in the wood and the whiskey. In the one full scale test of the control
system, whiskey quality was in fact lowered and the test was discontinued.
However, analysis of the test indicates that certain design and operating
changes may have eliminated the whiskey quality problems.
The cost problems discussed above and the failure of the full scale test
show that control of emissions from whiskey warehousing has not been demonstrated
at this time. However, the control systems show a potential for breaking
even or producing a profit, an unusual characteristic for a control system.
Even without credit for recovered alcohol, the control system costs 7-10<£/proof
gallon, which compares favorably to a production cost of $2.10/proof gallon.
In addition, engineering analysis indicates that problems with whiskey
quality can potentially be solved with proper design and operation. Thus, it
appears possible that further work could demonstrate the feasibility of
control. This work would include the following:
1. investigation of alternate carbon regeneration techniques, for example
electric heating/vacuum regeneration
2. additional economic analysis. A low sensitivity of liquor demand to
price changes and the large percentage of liquor prices made up by taxes may allow
the costs of the control to be passed on even without credit for recovered alcohol.
3. additional testing of the control systems
4. scheduled tests to demonstrate an alternate aging system. This system
is discussed in section 4.5.
This further work was not able to be completed at the publication date of
this document.
1-4
-------
2.0 WHISKEY WAREHOUSING AND AGING
The manufacture of whiskey involves two distinct steps - the production
of imaged whiskey from cereal grains and the maturation of this whiskey by
storage in charred white oak barrels.
In the production of unaged whiskey, grain is first milled, then cooked
in water to solubilize the starches. The solubilized starches are then mixed
with partially germinated grain. This step results in the starches being hydrolyzed
to sugars by the enzymes in the germinataeUgrain. The sugars are then fermented
with yeast and the resulting mixture is distilled to produce unaged whiskey.
The production of unaged whiskey is a source of only a small percent of the
volatile organic chemicals emitted in whiskey manufacture. The emissions from
this first step are described in Appendix A.
The unaged whiskey, colorless and pungent tasting, must be aged by storage
in charred oak barrels to produce an alcoholic beverage with the traditional
characteristics of whiskey. This step, whiskey aging, is the major source of
emissions in whiskey manufacture and will be the principal focus of the report.
This chapter will describe whiskey warehousing operations and the physical and
chemical processes that occur as whiskey ages. Chapter-3 will present emission
factors for whiskey warehousing and the basis of these emission factors, and
Chapter 4 will describe possible emission controls and their advantages
and disadvantages.
2-1
-------
2.1 BARRELING AND WAREHOUSING
To produce an alcoholic beverage with the traditional qualities of
whiskey, the unaged whiskey is stored in new, white oak barrels, whose
head and staves have been charred. The barrels are normally constructed
of 25 staves from 2 to 3 cm in thickness and charred for 30 to 50 seconds.
The barrels typically hold 190 liters and are approximately 89 cm tall and
54 cm diameter at the head.
During aging, the barrels are stored in large warehouses. There are
three types of warehouse desiqn: brick and masonry rack design; metal clad,
wood-frame rack design; and palletized design. Rack designs consist of
multi-level lattice structures made of wood or metal, on which the barrels
are tightly packed on their sides in long parallel rows and supported by
beams at the ends of the barrels. In rack design warehouses, there are commonly
three to six levels of barrels per floor and five to ten floors per warehouse.
Brick rack designs have concrete floors, roof, and brick exteriors, with windows
normally on each floor for ventilation. Metal clad rack designs have corrogated
or sheet metal exterior and roof which are attached to the interior wood lattice.
The wood lattice supports the barrels and provides the structural support for the
warehouse. In contrast to brick and masonry warehouses, where the concrete
floors block internal air circulation., metal clad warehouses are open
internally with ventilation provided by windows or ventilators at the top
and bottom of the structure. Palletized design warehouses are single story
structures with barrels stored upright on pallets, with 15 barrels a pallet.
Palletized designs require more land than rack designs, but reduce the labor
required to handle the barrels.
2-2
-------
The barrel capacity range of warehouses varies as a function of design:
40,000 to 100,000 for brick rack designs, 20,000 barrels or less for metal
clad rack designs, and up to 35,000 for palletized designs. The absence of
water sprinklers for fire protection in metal clad rack warehouses limits
their size for insurance reasons.
The total barrel capacity of a typical warehousing operation ranges from
200,000 to 600,000 barrels. Brick warehouses are generally used in urban areas
because of fire and building codes, and metal clad warehouses are generally used
in rural areas. Metal clad warehouses are placed 60 meters or more
apart for fire protection and thus a large storage facility with 30 warehouses
will cover up to 450 hectares. Other smaller rural facilities may be dispersed
because of hilly terrain or to place the warehouses in the optimum location for
aging. A listing of barrels stored in Kentucky distilleries is presented in
Appendix B.
2.2 MECHANISMS OF AGING
The main components of whiskey, ethanol and water, are relatively
insignificant factors in its flavor intensity and palatability. The distinctive
qualities of whiskey are due for the most part to the trace constituents,
called "cogeners," present in the beverage. These substances are generated in
part during fermentation, but the majority are added in the course of aging.
/
During aging these trace constituents are added to the whiskey by three
mechanisms:
1. extraction--of organic substances from the wood and their transfer
to the whiskey,
2. oxidation of the original substances and of the extracted wood
material, and
2-3
-------
3. reaction between various organic substances present in the liquid
to form new products.
The nature and changes in the concentration of these trace constituents are shown
in a comprehensive study of whiskey during maturation by Liebmann and Scherl
2
of Schenley Distillers. Their study covered an 8 year period and included
analysis of 469 barrels. Table 2-1 presents the statistical design of the
major variables of the study and Table 2-2 lists the characteristics of whiskey
at various maturation times. The main changes in physical and chemical characteristic;
of whiskey, occurring as a function of time are shown in Figure 2-1.
There are several points to note concerning changes in whiskey during
aging as observed in the Liebmann and Scherl study. The fixed acids, furfural,
solids, color, and tannins in whiskey are added entirely during aging. (The
small amounts present initially in tne whiskey sampled in the study were due to
the fact that some of the whiskey had been treated with oak chips before barreling.)
In contrast, there are significant quantities of esters and fusel oil and
lesser quantities of total acids and aldehydes present prior to aging. The
concentration changes for most constituents are essentially complete by three
years of aging; however, esters and solids continue to show significant increases
in concentration beyond that time. The increase in aldehydes, acids and esters,
oxidation and reaction products of alcohols, show the importance of chemical
reactions in aging. In examining the chemical changes it is important to note
that there are only rough relations between chemical analysis and quality,
i.e., taste and aroma of whiskey. It is necessary to rely on the human
senses of taste and smell to detect fine variations and thus evaluate the quality
of whiskey.
The precise sequence and interdependence of the mechanisms responsible
for aging are quite complex and not completely understood. However, the
following paragraphs describe in general the chemical and physical phenomena
responsible for aging. The description is purposely qualitative since the
-------
Table 2-1. STATISTICAL DATA OF WHISKEY MATURATION STUDY BY LIEBMANN AND SCHERL
Grain formula
Type
Bourbon
60* corn
40% sraa 11 grain
7 5 % co c n
25% small grain
80* corn
20* small grain
88* corn
12% small grain
Rye
51% rye
49% other grains
Ho.
84
43
151
112
79
469
*
18
9
32
24
17
100
Distillation
Type
Singled
Doubled
Ho.
82
387
469
t
17
83
100
Treatment
Type
Untreated
Oak chip-treated
Nuchar- treated
No.
255
54
160
469
t
54
12
34
100
Warehouse
Type
Rack (wood)
Concrete
No.
219
250
469
%
47
53
100
Storage
Location
Louisville, Ky .
Schenley, Pa.
Lexington, Ky.
Lawrenceburg, Ind ,
Frank f ort , Ky .
No.
128
114
64
91
72
469
%
27
24
14
19
16
100
1X3
CJ1
Table 2-2. CHARACTERISTICS OF AMERICAN WHISKIES AT VARIOUS AGES2
AS«
Vr.
1
1
3
t
a
•
7
•
Mon.
0
1
3
fl
12
IS
2t
CO
3fl
42
4
21 a
2Q.3
31.1
35. S
33. B
41.8
44.7
47. n
48.0
si. a
5V 6
S7.6
01,2
62.0
04.4
M.8
AM*-
hyJen
1,4
2 1
2.8
3 3
4.1
4.H
5.S
58
no
n.o
B.I
e t
a. 2
e.3
6.5
7.0
7.0
70
70
Viir-
lurml
0.2
.2
.5
.e
.7
.8
.8
.0
.8
0
.8
.7
.7
.8
.8
.B
.8
20
2.0
Fii'H
Oil
111
123
131
131
132
132
134
131)
m
137
138
Salidi
8.7
44.1
WV.6
87.7
111 1
127. a
137. 5
147 7
152.7
I'i7.7
ifl:> . n
Iflfl.O
173.0
174.2
181.5
18fl 0
108. S
198. B
209. e
Color
(Df-mity)
O.03i
O.ir.il
0.2f>:>
tl 213
O.JSi
0.31IS
0 32H
o :ui
0.3'ii
l> 310
o 3'r>
0 3H7
0 3C,S
o . :ir.j
0.3SO
0 3S'>
a, zm
0.413
0.449
• AH f turn rr^rmmt average ratuet and a« M pressed W grama par 100 liters at 10O proo
praeet) a o
w
53
I>H
.92
«a
4rt
.39
3*1
.2fl
.2ft
.28
.27
.29
2fl
.20
26
J(V
.24
24
.23
.32
.20
. eKcrfii proot (FK
-------
I
114
110
106
103
98
94
13
3 . 0
13
4 38 *8 00 73
An. Mouth*
Figure 1. Proof
timpttml mi UiaiU
84 98
12 24 30 48 60
Ac*. Month*
Figure 3. Fixed Acidity
72 84
too
8 M
t
J eo
8
is «o
^
t
1 -'..
J2 24 3fl 48 80 7*
A(«. Month*
Figure 2. Total Acids
^—A»*r«e»« diipmlwi llmili
§ 80
00
40
20
12 24 36 48 GO 72
A4«, Month*
Figure 4. Eaters
dUpmion limil*
•6
.. o
IS 34 36 48 60 72
Afc- Monthi
Figure S- Aldehyde*
84
|
•
§
o
II
24 afl 48 eo
A«*. Month*
. Furfural
T7
Figure 2-1. Effect of maturation on the physical and
chemical characteristics of whiskey,
Liebmann and Scherl study 2
2-6
-------
24
ART. Months
7. Fuacl Oil
30
i limit.
I
12
24 30 43 CO
AC=. Moaits
Fi(jurc 0. Solids
• A*ct*c«l • 4bpcnloa Unit*
72 S4
12 Si 39 43 00 72
Ace. Montti
Figure 9. Color (Density)
•op* Afwovi _^_ 4tefHW4flM UviitB
84 90
4.S
«.«
«**
f.2
4.0
a.s
12 24
38 41 M
AC«. Moothi
Figure 10. Tannins
>A**rm«et dl»|m«toa Unite
72 M W»
24 M 48 W) 72 B4
AM. Moathi
Figure 11. pH
' A»FT»gl I —
Figure 2-1. (cont.) Effect of maturation on the physical
and chemical characteristics of whiskey,
Liebmann and Scherl study^
2-7
-------
exact rates of the phenomena and the sensitivity of these phenomena to changes
in such variables as temperature and entry proof is not precisely
known.
The aging process begins when the barrel is filled with whiskey and the
charred wood becomes saturated with liquid. The liquid extracts from the charred
wood partially oxidized organic substances in the char, the biologically formed
organic substances in the uncharred wood, plus color and various solids.
This material is transferred to the bulk liquid in the barrel by simple
diffusion, by convection currents in the bulk liquid and by temperature cycling.
Temperature cycling causes transfer of material in the following way. As the
barrel heats up, the gas above the liquid increases in pressure and forces
liquid into the barrel wood. When the barrel cools and the gas pressure
^
drops, the liquid flows out of^wood into the bulk liquid, carrying wood constituents
with it. The materials transferred and originally in the wood react to form
new compounds. These reactions occur on the surface of the wood, with the
char acting as a catalyst, and in the bulk liquid. In addition, oxidation
of chemical substances occurjas a result of the slow diffusion of air into
the barrel liquid.
The rates of extraction, transfer, and reaction depend on temperature
and the concentrations of various whiskey constituents. The effect of temperature
is straightforward - higher temperatures increase the rates of extraction, transfer
by diffusion, and reaction. Also, temperature changes cause convection currents
in the liquid and pressure changes in the gas affecting transfer. The effect
of concentration is more complex. The rate of extraction of various char
and wood constituents will depend on the relative concentration of ethanol and
water in the wood, since the constituents will exhibit differing solubilities
in water vs. ethanol. The rate of extraction will also depend on the overall
2-8
-------
concentration of liquid in the wood. The rate of diffusion will depend on the
difference of concentrations of constituents in the wood, liquid, and
air around the barrel. The rates of reaction will increase or decrease with
the concentration of constituents.
The equilibrium concentrations of the various whiskey components depend
heavily on the air flow around the barrel. A large air flow will lower the
concentration of water, ethanol, and trace constituents in the air and increase
the concentration gradient between the air and the barrel wood. This will have a
number of effects. First, the larger concentration gradient will cause water
and ethanol to evaporate faster and the ethanol/water content of the barrel
2
wood to drop. An example of this phenomena is that,blotter strip whose end
is stuck in water will be drier and water will evaporate faster with air blowing
over it. The faster evaporating ethanol and water will draw more wood constituents
out than normal, allowing less to travel inward to the bulk liquid. Also the lower
liquid content of the wood will effect extraction. Finally, the larger concentration
gradient for trace constiuents will cause these substances to evaporate to the air
faster, again upsetting their inward transfer to the liquid. Figures 2-2 and 2-3
illustrate these various transfer mechanisms, and other aspects of aging.
2,3 WAREHOUSE OPERATION
The preceding discussion illustrates the importance of correctly controlling
the barrel environment to produce a whiskey of a desired quality. Since each
distiller desires to produce a whiskey with a quality distinctive to their
brand, the various distillers control the barrel environment differently by
operating their warehouses in different manners. However, it must be kept in
mind that the effects on whiskey quality of such warehouse parameters as
temperature, temperature cycling, humidity and ventilation are not precisely known.
2-9
-------
EVAPORATION OF
ETHANOL & WATER THROUGH
CAPILLARY ACTION
ATMOSPHERE
AIR
DIFFUSION '
EVAPORATION -^
OFETHANOL
& WATER THROUGH
BARREL WOOD
PRESSURE
CHANGES
IN GAS
LIQUID SURFACE
LIQUID FORCED INTO
WOOD BY PRESSURE CHANGES
REACTIONS BETWEEN ETHANOL,
s. TRACE CONSTITUENTS IN THE WOOD,
LIQUID AND AIR. THESE REACTIONS
OCCUR IN THE LIQUID AND AT THE SURFACE
OF THE WOOD.
DIFFUSION OF WOOD
CONSTITUENTS FROM CHAR
LAYER &UNCHARRED
WOOD
BARREL
STAVE
CONVECTION CURRENTS MIXING
CONSTITUENTS INTO THE
BULK LIQUID
Figure 2-2. Mechanisms of whiskey aging.
2-10
-------
ATMOSPHERE
W2
THE DIRECTION OF THE
ARROWS INDICATES THE
DIRECTION OF THE
DIFFUSION& THE SIZE
OF THE ARROWS THE
RELATIVE AMOUNTS
E-ETHANOL
W-WATER
TC-TRACE CONSTITUENTS
THE TRACE CONSTITUENTS TRAVEL IN THE
ETHANOL/WATER LIQUID. THUS, THE RATE
OF DIFFUSION OF TRACE CONSTITUENTS
DEPENDS NOT ONLY ON THEIR CONCENTRA-
TION GRADIENTS, BUT ALSO ON THE OVER-
ALL RATE OF DIFFUSION OF THE ETHANOL/
WATER MIXTURE IN THE BARREL WOOD.
Figure 2-3. Diffusion through barrel staves in whiskey aging.
2-11
-------
Thus, present methods of warehouse operation have not been developed by design and
calculation; rather, each distiller's operation is for the most part the result
of tradition and experience.
Other factors besides quality influence warehouse operation. These include
the differing construction costs between metal clad and brick designs, the energy
required if heating is used in the winter, the labor involved in moving barrels
and opening and closing windows, the level of evaporative losses, and the
savings in barrel costs if whiskey entry proof is increased.
The most important variation in warehouse operation is the type of warehouse:
brick, metal clad or palletized. One aging/quality philosophy is that the
best whiskey is produced when the barrel follows natural conditions during
aging. Thus, metal clad warehouses are used since their exteriors are
designed only to keep rain and snow from the barrels and provide no additional
protection from the weather. However, the labor savings involved in palletized
designs, construction costs and fire codes also influence the choice of
warehouse type.
Another area where variations in practice occur is the type of ventilation
provided for the solar heating effect. The large roof area of palletized
designs and the poor insulation characteristics of metal clad designs allow
relatively high rates of solar heat transfer through the roof and upper levels.
If no natural or forced air circulation is provided, a hot, stagnant air
./
mass develops in the upper area and a sizable temperature difference can
develop between the top and bottom of the warehouse. This effect is commonly
observed in metal clad warehouses during the summer, when temperatures of
120 to 140°F can develop in the top floor while temperatures at the bottom
are only 65 to 70°F.
2-12
-------
Various practices are followed with respect to this solar heating effect.
Some distillers desire the elevated temperatures to achieve the type of aging they
desire and thus close the bottom or top windows to create these high temperatures.
Others provide for ventilation at the top and bottom of the warehouse to
induce air flow and reduce the temperature difference. This is done not only
to produce different temperatures for aging, but also to reduce the high
evaporation losses at the elevated temperatures and to produce more uniform
aging conditions in the warehouse. One distiller, in an effort to achieve complete
uniformity of conditions and product, has sealed and insulated his metal
clad houses and installed a central ventilation and heating system.
Variations in operating methods also exist among brick warehouses
and between brick and metal clad houses. Brick houses have much better
insulation characteristics, and thus do not experience the extreme temperature
gradients in the warehouse during summer. Thus, whereas barrels stored in
metal clad houses are rotated to average out the exposure temperature
barrel rotation is not nearly as critical in brick warehouses.
The insulating characteristics of brick warehouses also allow for heating in
winter, whereas metal clads are allowed to follow the ambient temperature.
In addition, among brick warehouses, different heating practices are used.
Distillers not only maintain different temperatures in the winter, but also
practice different cycling techniques. Some have only seasonal cycles> cooling
in fall and warming in spring, while others intentionally increase and decrease
the warehouse temperature several times in winter to produce the type of
aging they desire. Variations between distillers also occur in the practice
of summer ventilation. Some simply open the windows, while two locations have
completely closed buildings and ventilate with fans.
2-13
-------
Other more detailed variations undoubtedly exist. These include the time
of the year windows are closed or heating starting, the length of temperature
cycling, the frequency windows are open and shut, and the humidity characteristics
of the spot selected for the warehouse. All of these variations illustrate the
number of differing aging philosophies and traditions. The practices of
several distillers are shown on Table 2-3.
2-14
-------
Table 2-3
Warehousing Operations
ick & Masonry Design
Forced Air
Company
A
A, Bldg. E
B
C
D
jtal Clad
Company
E
F
5 present
previously
H
Heating in
Winter
Yes
Yes
Yes
Yes
No
Heating
in Winter
No
No
No
No
No
Open Windows
in Summer
Yes
No, no windows
No
Yes
Yes
Windows open
in summer
Bottom Toj:
Yes Yes
No Yes
Yes Yes
No Yes
Yes No
Ventilation
in Summer
No
Yes
Yes
No
No
Barrel
Rotation
Temperature
Cycles
seasonal
seasonal
several times
in winter
several times
in winter
seasonal
Temperature
Summer Winter
Ambient
Ambient
Ambient
Ambient
Ambient
40°F
40°F
55°F
40°F
Ambient
Temperature - summer
Top Bottom
every 2 years 95°F
85°F
every 2 years 120°F
Not stated
Not stated
New barrels
Not Stated
120°F
elevated
65°F
70°F
started at top
and moved down
The warehouses have been sealed and
insulated and a central heating/
ventilation system installed
temperature cycling in winter;
in summer forced air
ventilation used to keep the
AT to a minimum
2-15
-------
2.4 REFERENCES
1. Liebmann, A, J., and B. Scherl. Changes in Whiskey While Maturing.
Industrial and Engineering Chemistry. 41:534-543, 1949.
2. Reference 1
3. Trip Report by Terry Briggs, Pedco, Cincinnati, Ohio on a visit to
Jim Beam Distillery, Clermont County, Kentucky, April 7, 1977.
4. Trip Report by Terry Briggs, Pedco, Cincinnati, Ohio on a visit
to Schenley Distillery, Louisville, Kentucky, April 7, 1977.
5. Trip Report by Terry Briggs, Pedco, Cincinnati, Ohio on a visit
to Barton Distillery, Bardstown, Kentucky, April 7, 1977.
6. Trip Report by Terry Briggs, Pedco, Cincinnati, Ohio on a visit
to Seagrams, Inc., Lawrenceburg, Indiana, March 30, 1977.
7. Trip Report by Terry Briggs, Pedco, Cincinnati, Ohio on a visit
to Seagrams, Inc., Louisville, Kentucky, March 30, 1977.
8. Trip Report by Terry Briggs, Pedco, Cincinnati, Ohio on a visit
to Brown-Foreman, Louisville, Kentucky, April 8, 1977.
9. Trip Report by Terry Briggs, Pedco, Cincinnati, Ohio, on a visit
to Heaven Hill Distillery, Bardstown, Kentucky, April 7, 1977.
10. Telephone conversation between Mr. T. W. Samuels, Jr. President,
Maker's Mark Distillery, Inc., Louisville, Kentucky and David C. Mascone,
U. S. EPA, September 19, 1977.
11. Telephone conversation between Mr. David C. Mascone, U. S. EPA and
Bill Padgett, Austin Nichols, Lawrenceburg, Kentucky, January 19, 1978,
2-16
-------
3,0 VOLATILE ORGANIC EMISSIONS FROM
WHISKEY WAREHOUSING
This chapter will describe the volatile organic emissions from whiskey
warehousing, develop an emission factor for these emissions and present an
estimated national emission inventory.
3.1 EMISSION SOURCE DESCRIPTION
The two sources of ethanol in whiskey warehousing are evaporation from
the barrel wood during storage and evaporation from the saturated wood after
the barrel is emptied. These emission sources are described below.
The first emission, evaporation during storage, occurs when liquid
diffuses through the barrel staves and heads via the wood pores or travels
by capillary action to the ends of the barrel staves. The liquid evaporated
is both water and ethanol, with minor amounts of trace constituents. As
discussed in Chapter 2.0, this ability of the barrel to "breath", i.e. allow
liquid to evaporate and air to enter, is important to aging. Attempts made to
age whiskey in sealed containers and thus prevent losses have proven unsuccessful
since little aging occurred.
The rate of evaporation during aging is not constant. During the first
six months to a year, the evaporation rate is low, since the wood starts dry
and must become saturated before evaporation occurs. After saturation, the
evaporation rate is greatest but decreases as the evaporation lowers the liquid
level in the barrel. The lower liquid level decreases the surface area of the
liquid in contact with the wood and thus the surface area subject to evaporation.
3-1
-------
The second emission, evaporation after barrel emptying, occurs when
the saturated barrels are stored after emptying. The amount and location of
these emissions depend on the use that the distillers find for the barrels.
A significant fraction are stored outside for lengthy periods during which
much of the alcohol evaporates. Even if further use is found for the barrels,
the bound alcohol will still evaporate if the barrels are stored long enough
before reuse. Potential end uses for used barrels are aging Scotch, Canadian
whiskies and American light whiskies, and as fuel or for decorative purposes.
Federal law prohibits the use of used barrels in bourbon and American blended
whiskey.
3.2 WHISKEY WAREHOUSING EMISSION FACTORS
Two sources of data are available to develop emissions factors for whiskey
warehousing - aggregate loss data from IRS publications and individual loss
data from specific distillers.
3.2.1 Emission Factors from IRS Data
The aggregate loss data from IRS publications are presented in
1 2
Table 3-1. ' Shown on this table are data on whiskey withdrawals, losses and
stocks for 1974, 1975, and 1976, along with emission factors calculated from
>£
this data. Withdrawls represent whiskey removed from storage for consumption.
Losses represent the difference between the original and withdrawn amounts, i.e.
that amount of whiskey lost due to evaporation and barrel soakage, plus theft,
spills, etc. Average stocks represent an average of the amount of whiskey held
in storage for that year and the previous five.
Three emission factors were developed from this data. Emission Factor I
represents the fraction of whiskey production lost and equals .2 proof gallons
lost for each proof gallon whiskey produced. This factor was computed by dividing
3-2
-------
Table 3-1. LOSSES, WITHDRAWALS, AND STOCKS OF WHISKEY FOR THE U.S.
Column
~v
1
i Year
1976
1975
1974
2
Withdrawals
134.8
136.9
138.1
3
Losses
33.7
36.0
33.9
4
Withdrawals
+ Losses i
168.5
172.9
172.0
5
Emission
Factor I ,
.200
.208
.197
6
2
Average
iStocks
870.6
910.0
935.7
7
Emission
Factor II
.039 ™rf
.039 '*<*
.036 ?i-
8
Emission f
Factor III1
3.2
3.0
'4-
Computed by dividing column 3 by column 4, represents pg lost/pg whiskey produced.
"Represents the average of the stocks of whiskey in storage for the previous 6 years.
Computed by dividing column 3 by column 6, represents (pg lost/year)/pg whiskey in storage.
i
Computed by multiplying column 7 by 55 pg/barrel and 1.5 kg/pg lost, represents kg ethanol lost/barrel-yr.
Table 3-2. BARREL" SOAKAGE LOSSES
Source
Brown-Foreman
Boruff & Rittschof
Gallagher, et. al.
Schenley
Barrel Soakage
kcL liquid Ibs liquid
7.3 6,0
10.3 io. 0
8.6 ^;,o
5.5 ' •
11.4 l» "1
16
22.6
19
12
25
Aging Time
years
5
8
5
1
10
1
Best Fit Equation Mo. of years
kg liquid soakage
(i ,e. water + ethanol )
-.67(aging time,yrs) +4.7
for years 1 & greater
5
8
5
1
10
kg lost-equation
8.1
10.0
8.1
5.4
11.4
-------
total losses by total production (losses plus withdrawals). Emission Factor
II represents the loss rate based on stored whiskey and equals .038 proof
gallons lost for each proof gallon in storage each year. This factor was
computed by dividing total losses by average stocks. The number of proof
gallons in stock was taken to be the average of the number of proof gallons
i
in stock for that year and the previous five. The 6-year average stock
was used since losses recorded for a given year represent losses on barrels
emptied that year. These losses actually occurred not only during that year,
but in previous years while the barrel was in storage. Six years is an
approximation of the period of barrel storage - some of the losses for a
given year come from barrels stored eight years and more, whereas some
stored six years ago have already been emptied for four year old whiskey.
Emission Factor III represents a weight loss rate per barrel per year and equals
3.2 kg ethanol/per barrel each year, This factor was computed by multiplying
Emission Factor II by 55 proof gallons per barrel and 1.5 kg ethanol per
proof gallon. It is important to note that the above figures include losses
for both evaporation during storage and soaking into the barrel.
3.2.2 Emission Factors from Individual Distiller Data
The loss rate data from individual distillers and from experiments cover
two areas, barrel soakage losses and evaporation losses during storage. These
are discussed below.
The data available on barrel soakage losses are presented in Table 3-2. * ' >l:i
The table shows the available data on total liquid soakage vs. aging time,
plus a best fit equation for this data. The table indicates a rapid saturation
of the barrel during the first year, followed by a constant, but slow, increase
in weight during subsequent years. It should be noted that the data are for
liquid soakage, i.e., both water and ethanol. Work by Boruff and Rittschof indicates
that the proof of the liquid in the barrel wood is approximately the same as
3-4
-------
the proof of the stored whiskey; this permits a conversion from kg liquid to
kg ethanol. Thus, a typical barrel storing 120 proof whiskey emptied after
four years contains 3.8 kg of ethanol in the saturated wood.
The data from experiments and individual distillers on evaporation during
storage are shown on Table 3-3. " The cumulative loss represents the total
ethanol loss due to evaporation during the aging time shown. The annualized
loss rate expresses this total at a constant yearly loss rate and was computed
by dividing the cumulative loss by the aging time. Table 3-3 also shows a
best fit equation for annualized losses for aging times of four years or more.
Annualized loss rates vs. aging time, as computed from the data and equation
in Table 3-3, are shown on Table 3-4. Also shown on Table 3-4 are computed
cumulative loss and computed incremental loss. Cumulative loss was calculated by
multiplying the aging time by the annualized loss rates from the best fit equation. •
Incremental loss was computed by subtracting the computed cumulative loss for two
successive years. This later number represents the additional evaporative loss
during the given year of aging.
Figure 3-1 shows graphically the data on annualized loss rate from Table 3-3
and the computed annualized and incremental loss rates from Table 3-4. The
graph clearly shows the wide variation in evaporative loss between distillers.
These variations can be explained qualitatively by variations between distillers
in such warehouse parameters as temperature, ventilation patters and temperature
cycling. However, because of the large number of conditions that affect evaporation
and the limited knowledge on the precise effects of the conditions on the rate of
evaporation, no attempt was made to statistically relate warehouse conditions
to evaporative loss.
Figure 3-1 also shows the variation in the incremental loss rate during
aging, with the rate increasing during the first two years and decreasing in
3-5
-------
Table 3-3. EVAPORATIVE LOSSES DURING STORAGE
Source No.
Gallagher, et. al.
Gallagher, et. al.
A
C
E
F
C
Boruff & Rittschof
F
I
Aging Time
Years
1
2
4
4
A
5
6
8
9
10
Cumulative Loss
kg ethanol /barrel
2.35
6.59
9.52
15.60
9.32
14.45
20.88
17.76
18.81
26.70
Annualized loss
kg ethanol/barrel-yr
2.35
3.30
2.38,
3.90
2.33
2.89
3.48
2.22
2.09
'2.67
O
Best fit Equatiort-Annualized Loss
For years 4 & greater
Annualized Loss (kg ethanol/barrel-yr)
= -.101(aging Time, yrs) +3.38
Letters indicate data from individual distillers; Letters refer back to same distillers as Table/2-3
DAnnualized losses assuming equal loss each year.
Table 3-4. COMPUTED ANNUALIZED, CUMULATIVE & INCREMENTAL LOSSES
Aging Time
Years
1
2
3
4
5
6
7
8
9
10
i ' •£-
Annualized Loss kg/barrel}- yra
* 2.35 •
-v 3.30
3.10
2.98
2.88
2.78
2.67
2.57
2.47
2.37
Cumulative Loss kg/barrel
2.35 y,
6.60 ^
9.30 >3
11.92 '/„
14.40 '/
16.68
18.69
20.56
22.23
23.70
Incremental Loss kg/barrel-yrc
2.35
4.25
2.70
2.62
2.48
2.28
2.01
1.87
1.67
1.47
'i j '
Years 1 & 2 are taken from Gallagher, et. al.; years 3 & greater from the best fit equation, Table 3-3.
"'Annualized loss times aging time.
"Difference between cumulative loss for successive years.
-------
o
35
MEASURED ANNUALIZED LOSS RATES
CALCULATED INCREMENTAL LOSS RATE
BEST-FIT FOR ANNUALIZED LOSS RATE
456
WHISKEY AGE, years
Figure 3-1. Emission rate relationships in the whiskey aging process.
3-7
-------
subsequent years. This is in agreement with the theory discussed early.
This variation in the incremental loss rate means that the age mix of the
barrels in storage will affect the emission rate. Since barrels of different
age have different evaporative loss rates, the total emissions will be
determined by the fraction of barrels at each age.
Three different barrel age distributions were used to calculated emission
factors: (1) the age distribution of bonded whiskey in Kentucky at the end of
1975; (2) an age distribution based on fluctuating market from year to year;
and (3) the age distribution based on distillers producing mainly four year
old whiskey. Table 3-5 presents the barrel age distribution for the three
cases and the respective emission factors of 2.55 kg/barrel-yr for case one,
2.74 kg/barrel-yr for case two, and 2.89 kg/barrel-yr for case three. These
emission factors were calculated by multiplying the fraction of the barrels at
a given age by the incremental loss for that age in Tab "he 3-5. The four distillers
producing primarily four and six year old whiskey used in case three are
Jim Beam, Clermont, Kentucky; Jim Beam, Beam, Kentucky; Brown-Foreman, Louisville,
Kentucky; and Fleischmann, Owensboro, Kentucky.
The above emission factors represent evaporative losses during storage only.
To determine overall emission factors, losses due to barrel soakage must be
included. This loss is computed by assuming that the number of barrels emptied
in a year equals the number of barrels one year old, and that the average barrel
has a soakage equivalent to a five year old barrel. This figure is 4.2 kg ethanol/
barrel. The overall emission factor is therefore:
Aging + Soakage = Total,Emissions
case one) 2.55 +,4.1, (.112) = 3.02 '" kg/barrel-yr
case three) 2.39 + 4.2 (.181) ~ 3.65 •t.s'"'
In the preceding discussion, the variations in evaporative loss rate
during aging were averaged together to develop a single emission factor.
3-f
-------
Table 3-5. WAREHOUSE BARREL AGE DISTRIBUTION
(1) Whiskey by Various Periods of Production Remaining in
Bondea Warehouses in Kentucky as of Dec. 31, 1975.
Age
0-1
1-2
2-3
3-4
4-5
5-6
6-7
7-8
9 +
Barrels in bond
in Kentucky
685,600
657,600
813,800
943,400
868,700
821,000
761,900
349,600
247,200
6,148,600
Fraction
by year
0.112
0.107
0.132
0.153
0.141
0.134
[J r j r ^ "
•T • ' -^ ;. ' —
J . --*• .
fj-^y;'-^
0 roof- Average barrel loss
'',,2.55 kg/barrel-year
^'''o4-<
0.124 "}^.o-
0.057
0.040
1.000
0./Q6J
2i^ii -
.»•'' " "-" '^ 'r i '
(2) Barrel Age Distribution Assuming a Uniform Year-to-Year
Consumption Rate (100 bbl/yr basis)
Age
0-1
1-2
2-3
3-4
4-5
5-6
6-7
7-8
9+
%
Used
(end of year)
35
20
15
20
10
Total
by year
100
100
100
100
65
45
30
30
10
580
Fraction in
warehouse
by year
0.172
0.172
0.172
0.172
0.112
0.079
0.052
0.052
0.017
1.000
Average barrel loss
2.74 kg/barrel-year
(3) 4 to 6 yr Whiskey Production
Age
0-1
1-2
2-3
3-4
4-5
5-6
6-7
7-8
9+
Beam
Beam, Ky.
58948
64014
98247
91239
17572
1110
303
2122
5698
Bean
Clermont, Ky.
60743
74076
78559
84464
24102
31594
14981
25207
12069
Brown-Forman
Louisville, Ky.
97000
104437
41840
63371
60514
37320
4321
2783
858
Fleishmann
Owensboro, Ky.
30901
38568
35413
36411
30412
35963
5412
208
Overall age
distribution
0.181
0.205
0.185
0.201
0.097
0.077
0.018
0.022
0.014
1.000
Average barrel loss
2,74 kg/barrel-vear
3-9
-------
This single emission factor was then used together with data on barrel age
distributions to compute several emission factors. A second method of
developing emission factors from the loss data reported by individual distillers
is to group the data into higher and lower measured annualized loss rates.
As noted previously in Chapter 3, large variations in measured annualized loss
rate result from differing warehouse operations. The analysis of the loss rates
by dividing them into higher and lower values will provide two emission factors
characterizing the spread of emissions caused by differences in warehouse
operations. Examination of Figure 3-1 shows that the bottom four and top
three data points for measured annualized loss fit into two convenient groups.
Analysis of these groups results in emission factors of 2.3 and 3.6 kg/barrel-yr
for evaporative loss during aging.
It should be noted that the above analysis was not performed rigorously.
A rigorous analysis would require that the annualized loss data be converted
to incremental losses, and then the incremental loss applied to barrel age
distributions. This was not done because it was felt that three data points
(four in the lower value case) were not sufficient for these conversions to remain
statistically meaningful. Thus, the emission factors of 2.3 and 3.6 kg/barrel-yr
were determined by drawing lines, lines through the bottom four and top three
points for measured annualized losses (Figure 3-1) and the loss rate at year
five were taken to be the appropriate emission factor.
All the emission factors for volatile organic chemicals from whiskey
warehousing are summarized in Table 3-6. The emission factors based on the
variations in warehouse operations are used in designing and costing the
control system. The emission factors developed from the barrel age distributions,
along with Emission Factor III from the IRS data, are used to develop emission
inventories. Finally, Emission Factor I from the IRS data is used to relate
3-10
-------
Table 3-6. SUMMARY OF EMISSION FACTORS
WHISKEY WAREHOUSING
Source
Figure
Description
IRS Publication
,3-4
Individual Distiller
Data & Experiments
JO
I
.20 proof gallons lost/proof gallons produced*
.038 proof gallons lost/proof gallons storage-yr*
j--'
3.2 kg ethanol/barrel-yr*
3.8 kg ethanol soakage/barrel
3.02,3.46,3.65 kg ethanol/barrel-year
2.3,3.6 kg ethanol/barrel-yr
represents fraction of production lost
represents fraction of storage lost per
year
represents amount of ethanol lost per
barrel in storage per year
represents amount of ethanol lost per
barrel due to sjaakagj into wood. The
figure is for aTarrel stored 4 years.
represents amount of ethanol lost due
to both evaporation during storage and
soakage for various barrel age
distributions
represents the range of ethanol loss durir
storage caused by differing methods of
warehouse operation; does not include
soakage loss
*These figures include all types of loss - evaporation during storage, soakage into the barrel, plus leakage, theft,etc.
-------
whiskey sales to markets in the discussion of reuse of the recovered alcohol.
The reason for using each emission factor for the uses described above is given
with the calculations involving that emission factor.
3.3 EMISSION INVENTORY
Total emission estimates are developed for three areas: (1) typical size
distilleries, (2) States; and (3) nationwide.
Two representative facilities were chosen to develop emission totals for
typical size distilleries: (1) a large 400,000 barrel facility producing primarily
four year whiskies and (2) a smaller 50,000 barrel facility producing whiskies
up to eight years and older. To compute the emission total for the 400,000
barrel facility the emission factor used is that of case three in on page 3-9
This emission factor is used since the barrel age distribution for case three
and for the 400,000 barrel facility are both based on producing four year old
whiskies. For the 50,000 barrel facility, the emission factor used is that
of case one on page 3-9. This emission factor is used since the Kentucky
barrel age distribution approximates those of distillers producing eight year
and older whiskies. The emission totals for the large distillery is 400,000
barrels x 3.65 kg/barrel-yr = 1460 MT/yr and for the large distillery 50,000
barrels x 3.02 kg/barrel-yr = 151 MT/yr.
Total emission estimates will be developed for five States - Kentucky,
Indiana, Illinois, Tennessee, and Maryland. Table 3-7 shows the number of
barrels stored in each State and the total emission estimate. The emission
factor used was 3.2 kg/barrel year, based on the aggregate loss data from IRS
publications. This emission factor was used since, being based on the widest ,
3-12
-------
Table 3-7, TOTAL EMISSION ESTIMATE BY STATE
No. of Barrels
in Storage Total Emissions
State June, 1976, Thousands (MT/yr)
Kentucky
Illinois
Indiana
Maryland
Tennessee
6130
1290
2260
640
580
19,620
4,130
7,240
2,050
1,780
data base, it was most likely to have correctly averaged the variation in barrel
emission rates that occur between warehouses.
The national emission total estimate is 38,170 MT/yr, based on 11.9 million
barrels stored in June, 1976. The five States above represent 91 percent
of the estimated emissions.
3-13
-------
3.4 REFERENCES
1. U. S. Department of the Treasury, Bureau of Alcohol, Tobacco and
Firearms. Alcohol, Tobacco and Firearms - Summary Statistics, FY 1975.
ATF P 1323.1 (4-76), U.S. Government Printing Office, Washington, D.C.
2. U. S. Department of the Treasury, Bureau of Alcohol, Tobacco and
Firearms. Alcohol, Tobacco and Firearms - Summary Statistics, FY 1976
ATF P 1323.1 (4-77), U. S. Government Printing Office, Washington, D.C.
3. Trip Report by Terry Briggs, Pedco, Cincinnati, Ohio on a visit
to Brown-Foreman, Louisville, Kentucky, April 8, 1977.
4. Gallagher, M., P. Kolachov, and H. F. Willkie. Whiskey Losses
During Aging. Industry and Engineering Chemistry 34:992-995, 1942.
5. Boruff, C. S., and L. A. Rittschof. Effects of Barreling Proof on
the Aging of American Whiskies. Agricultural and Food Chemistry,
7(9): 630-633, 1959.
6. Trip Report by Terry Briggs, Pedco, Cincinnati, Ohio on a visit to
Schenley Distillery, Louisville, Kentucky, April 7, 1977.
7. Reference 4.
8. Reference 5.
9. Reference 6.
10. Trip Report by Terry Briggs, Pedco, Cincinnati, Ohio on a visit to
Seagrams, Inc., Lawrenceburg, Indiana, March 30, 1977.
11. Letter from Dr. Alan T. Thomas, Brown-Forman Distillers, Louisville, Kentucky
to Pedco, Cincinnati, Ohio, April 19, 1977.
12. Trip Report by Terry Briggs, Pedco, Cincinnati, Ohio on a visit to
Jim Beam Distillery, Clermont County, Kentucky, April 7, 1977.
13. Trip Report by Terry Briggs, Pedco, Cincinnati, Ohio on a visit
to Barton Distillery, Bardstown, Kentucky, April 7, 1977,
3-14
-------
14. Liberty National Bank and Trust Co, of Louisville, Kentucky. Distilled
Spirits in Bonded Warehouses in Kentucky on December 31, 1975.
Bulletin 114, Louisville, Kentucky.
15. Reference 14.
16. Reference 2.
3-15
-------
-------
4.0 WAREHOUSE EMISSION CONTROL
Two methods for reduction of warehouse emissions were investigated:
1) carbon adsorption (CA) and 2) an-alternate aging system. The second method of
control is in early development and will require a number of years for testing.
However, the system's potential for large reduction in aging costs makes it
attractive as a control method, given successful testing.
4.1 CARBON ADSORPTION - SYSTEM DESCRIPTION
Controlling warehouse emissions by carbon adsorption would involve
closing the warehouse and ducting the interior to a carbon adsorption unit.
For brick warehouses, this would involve shutting most windows, doors, and
ventilators, leaving some open for intake air, and running ductwork along the
exterior of the building to the various floors. In some metal clad warehouses,
extra work may be required to close gaps between metal sheets, and between the
roof and the sides. However, most metal clad warehouses are tight enough in
construction that closing windows, doors, and ventilators would be sufficient.
The areas of sheet metal overlap would not need to be sealed since these areas
would provide the infiltration required to balance the air removed by the CA unit.
The CA unit itself would be a skid-mounted package system containing two
beds, fans, switching mechanisms and control, condenser/decanter, and internal
piping for steam and air flow. The unit would run on a two cycle system with
one bed adsorbing as the second was regenerated and cooled.
4-1
-------
4.2 CARBON ADSORPTION - COST ANALYSIS
In determining the costs of the carbon adsorption system, a number of
assumptions were made. These assumptions are listed in the sample
calculation shown later. Several of the major assumptions are discussed below.
First, two warehouse ethanol concentrations, 750 and 1500 ppm, were chosen.
The ethanol concentration must be stipulated since this parameter establishes
the flow rate of the CA unit. The 750 ppm level complies with the OSHA exposure
standard of 1000 ppm, 8 hour time-weighted average; the 1500 ppm level reflects
the concentration believed to be required for proper whiskey aging. (A more
complete discussion of the OSHA standard, whiskey quality and other impacts
of the control system is presented later.) Second, a range of installed costs
1234
vs. adsorber size was chosen based on the evaluation of a number of sources. ' ' '
The costs used ($20/scfm for units less than 4000 scfm, $!4/scfm for units
greater than 15,000 scfm, and $17 for those in between) represent figures in
the middle of the range presented by the sources. Third, a value of
$0.53/proof gallon of recovered alcohol was chosen. This was based on the
current price of 190 proof alcohol of $1.12/gallon (or $0.59/proof gallon)
discounted $0.04/proof gallon for transportation and $0.02/proof gallon for the
utilities required for redistillation of the recovered alcohol. Fourth,
85 percent recovery efficiency and an adsorber flow capacity of one and a half
times that based on a warehouse mass balance were chosen. The 85 percent recovery
allows for the maximum ethanol losses through openings in the warehouse,
through design of CA unit to achieve proper aging and during redistillation.
It is expected that greater efficiencies could be attained in many cases. The
1.5 times the mass balance design allows for variations in the adsorber air flow
rate required for proper whiskey aging and for recovery of the higher emissions
in summer caused by warmer temperatures. Finally, two barrel emission rates,
4-2
-------
2.3 and 3.6 kg/barrel-year, were chosen to examine the effect the variations
in emission rates caused by differing Warehouse operations have on system
design and cost. A sample calculation follows.
4-3
-------
Sample Calculation
1) Assumptions
^ /
- barrel emission rate of either 2.3 or 3.6 kg/barrel-yr. (Approximately
5.0 or 8.0 Ibs/barrel-yr) and warehouse ethanol concentration of either 750 or
1500 ppm.
- total installed costs (TIC)
$20/scfm for units^4000 scfm
$17/scfm 4000 scfm iunit 115,000 scfm
$14/scfm for units >15,000 scfm
- other costs
Annualized capital costs = 15 percent TIC
Taxes, insurance, etc = 4 percent TIC
Steam = 174/100 Ibs
Carbon = $1.00/lb
Electricity = 3£/kw.hr
Maintenance = .1 hr/hr operation at SlO/hr
- design will be based on yearly operation, with an overall 85 percent recovery,..
with the actual unit at 1.5x the calculated flow rate
- bed design parameters - two foot bed depth, operating velocity at 75 fpm,
7 in. HoO pressure drop, bed length 3 times bed width, 7 year bed life
- recovery parameters - bed capacity at 71bs ethanol/100 Ibs carbon, 3 Ibs steam/
Ib ethanol recovered, S0.53/pg ethanol recovered
2) Calculations
Example - 50,000 barrel warehouse, 750 ppm, 3.64 kg/barrel-yr (8.0 Ibs/barrel-yr)
- Mass Balance - the system must be designed so that the emission rate of
ethanol matches the removal rate by the CA unit.
emission rate = (No. of barrels)(Ibs/barrel-year)
removal rate = (scfm)ppm/106 (l/360)lb-mole/ft'j X
(46 lb/lb-mole)5.18(10)5 min/yr
_o
or (No. of barrels)(Ibs/barrel-yr) = scfm(ppm)6.62 (10)
thus (50,000)8 = scfm (750)6.62(10)~2
scfm = 8060
- Total Installed Costs
Unit size = 1.5(8060) = 12,090 scfm
$17/scfm (12,090) = $205,530
Annualized .15(5205,530) = $30,829
4-4
-------
- Other Costs
the amount of ethanol recovered =
.85(50,000)8 -
340,000 Ibs whiskey/yr
steam requirement =
340,000(3) = 1.02(10)° Ibs steam/yr
1.02 (10)6 $.17/100 Ibs steam =
$1734/yr
taxes, insurance, etc. =
.04 (TIC) = .04 ($205,530)
$8221
electricity =
(7 in
5.18 (
(7 in hLO) 249 pascals/in H70 = 1160 joules/nT Air ,
ID)5 min/yr (scfm) 1/35.3 (m3/ft3) = 1.47(10)^ (scfm)
using a 60 percent efficiency factor and 3.6 (10) joules/kw-hr
(7.06/.6) $.03/kwhr (8060) =
$2850/yr
maintenance and labor
.1 hr/hr operation x $10/hr =
8640 (.1) $10 = $8640
- Bed Design
scfm/linear velocity = surface area (SA)
SA - 12,090/75 = 161 ft2
L = 3U; SA = LW; SA = 3W2; W =
W = /r6773 = 7.3 ft
L = 3W = 22ft
Bed volume = 2 ft(SA) = 322
322 (30 Ibs/ft3) = 9660 Ibs/carbon
9660/7 yr ($l/lb) = $1380/yr Replacement carbon
Cycle time (assume 50 percent of ethanol removed from bed each cycle)
340,000 Ibs ethanol -yr/8640 = 39.4 Ibs/hr
9660 Ibs carbon (.07 Ibs ethanol/lb carbon). 5 removal efficiency =
338 Ibs recovered/cycle
338/39.3 = 8.5 hours
- Value of Recovered Alcohol
3.31 Ibs/pg
340,000/3.31 = 102,720 pg/yr
102,720 (.53) = S54,400/yr
4-5
-------
A comparison of six recovery system design cases is presented in
Table 4-1. The cases cover three warehouse sizes and two emission rate/warehouse
ethanol concentration combinations. The warehouse capacities chosen were 20,000,
50,000, and 100,000 barrels and represent typical sizes for existing metal clad
and brick units. The emission rate/warehouse ethanol concentrations chosen were
8 Ib/yr-barrel, 1500 ppm, and 5 Ib/yr-barrel 750 ppm. These cases represent the
highest and lowest net return rates, respectively.
The cost analysis as presented in Table 4-1 indicates that the control
system is financially feasible. Four of the six design cases offer net returns,
the remaining cases small net costs. When these net costs are calculated on a
per original proof gallon basis, aged 4 years, the cost is 0.52
-------
Table 4-1
Recovery System Costs
Case
No. of Barrels
Warehouse ethanol cone.,
Emission rate, Ibs/yr-barrel
Actual SCFM
Design, 1.5 Actual
Total Installed Costs
Annualized TIC
(TIC)
Whiskey recovered, Ibs/yr
Steam, 10^ Ibs/yr
Steam, $/yr
Electricity, $/yr
Tax, etc., $/yr
Maintenance, $/yr
SA, ft.2
Length, ft.
Width, ft.
Cycle Time, hrs.
Carbon, Ibs.
Carbin, $/yr
Proof gallon whiskey/yr
Whiskey value, $/yr
Total Annual Costs, $
New Cost (Return)
Cost/4 yr. Proof gal.
A
50,000
750
5
5,040
7,560
$128,520
$ 19,280
212,500
.637
$ 1,080
$ 1,780
$ 5,140
$ 8,640
100
17
5.8
8.5
12,000
$ 1,720
64,200
$ 34,030
$ 37,640
$ 3,610
.52
-------
4.3.1 Effect on Whiskey Quality
Whiskey quality is a critical factor in the marketability of whiskey
and in the distinction between the various brands. Alterations in whiskey
quality, i.e., taste and aroma, are a serious concern to distillers since
such alterations could affect consumer acceptance of the product and thus
reduce sales.
As discussed in Chapter 2, the taste and aroma qualities of whiskey are
largely a product of whiskey aging. Whiskey aging, in turn, is a complex
process composed of a number of interrelated chemical and physical mechanisms.
A CA system,with the potential for changing such warehouse conditions
as temperature, ventilation patterns, and humidity, could affect these aging
mechanisms and thus alter quality.
The installation and operation of a CA system could affect whiskey
quality in a number of ways. First, the increased ventilation provided by
a carbon adsorber could lower the concentration of ethanol, water and trace
constituents in the air around the barrel. This would increase the rates of
evaporation of these constituents and alter the liquid content of the wood,
upsetting the equilibrium concentrations in the wood, liquid and air and
potentially affecting quality.
Proper design of the CA system could eliminate this effect. If the flow
rate of the CA unit was adjusted so that the removal rate of air matched that
provided by natural ventilation, the ethanol, humidity and trace constituent
levels in the warehouse would remain unchanged. Since the CA unit is removing
air, and thus the components in the air, at the same rate as natural ventilation,
both natural ventilation and the CA system would provide for the same build up
of these components in the warehouse.
4-8
-------
However, other effects could occur. A CA unit provides a
continuous flow of air across the barrels; natural ventilation would be
intermittent. Thus, a CA unit would provide constant concentrations
around the barrels, whereas natural ventilation would allow the buildup
of stagnant layers. These stagnant layers would be removed occasionally
by the natural ventilation, producing a stop-start effect in which evaporation
occurs quickly after a draft and slows as the stagnant layer builds up.
Another effect would be the lowering of the temperature differentials
between the top and bottom of the warehouse. A CA would take air from several
floors within the warehouse and either recirculate this air or draw in new air
This mixing and ventilation would remove the hot, stagnant air at the top
of the warehouse, reducing the temperature on these floors. ,.
It appears that proper design could also eliminate these effects. The
proper stagnation periods and concentration levels could be maintained around the
barrel by adjusting the air flow rate and sequencing the ventilation. In such a
system, only two or three of the warehouse floors would be ducted to the carbon
adsorber at one time. Time-controlled dampers in the air exhaust lines
would sequence which floors received ventilation. During the period a floor
was off ventilation, the stagnation layers could build up. Elevated
temperatures at the top of the warehouse could be achieved by using very low
or no ventilation on the lower floors. Alternately, the system could be designed
to draw air upward through the warehouse. The air drawn in at the bottom would
be heated by the sun during the period it rose upward. Thus it appears that
the proper combination of air flow rates, ventilation patterns, air recirculation,
and other design parameters could reproduce most warehouse conditions. In
addition, it appears that this could be achieved in most cases with straight-
forward engineering and at moderate cost.
4-9
-------
However, proper design is not the only criterion; it is important to
know what conditions to reproduce. Given the complex nature of whiskey
aging, it is difficult to state precisely what are the conditions for proper
aging and thus how to design the CA system. This is especially true considering
the number of different brands of whiskey. Development of the system through
experimentation is also difficult. A minimum of 2 years is required to notice
quality changes in aging whiskey and 4 to 8 years to make a complete assessment.
Potentially, 2 or 3 four to eight year aging cycles could be required to adjust
the CA system to eliminate whiskey quality problems. Thus, the CA system's
affect on whiskey quality is indeterminate. It would appear possible to
design a system to reproduce the desired conditions but not possible to
state with precision what these conditions are.
4.3.2 Re-use of Recovered Alcohol
Important to the costs of the CA system is the ability to re-use the
recovered ethanol. This ability depends on two factors, the feasibility
and costs of converting the recovered ethanol to a product suitable for
use and the availability of markets for this converted product.
There are no market barriers to the re-use of the recovered alcohol,
once it has been converted to grain neutral spirits. Though tax regulations
prohibit its use in whiskies, the grain neutral spirits could be used in
7 8
vodka and gin, or denatured for chemical use. Consumption figures ' for
both these indicate that sufficient markets exist to absorb the recovered
product. If ethanol losses amount to 25 percent of the sales of American blended and
straight whiskies, this would provide 23 x 10° wine gallons/year or (assuming 100 prool
^Emission Factor II from the IRS data is .2 pg lost/pg produced. To calculate an
emission factor based on consumption, the losses must be subtracted from production
to arrive at a consumption figure, the loss rate on consumption is thus
.2/(l-.2) = .25
4-10
-------
whiskey) 15 x 106 190 proof gallons/year. The use of ethanol for gin and
vodka (assuming 100 proof for these products) is 53 x 1C 190 proof gallons/
year. Thus, the available market, gin, vodka, and industrial use, is 253 x 10
190 proof gallons/year (See Table 4-2). The recovered ethanol represents
11 percent of this market.
The conversion of the recovered ethanol to grain neutral spirits presents
no technical problems. The recovered alcohol is of sufficient quality for
distillation to grain spirits and the equipment and"procedures" to"perform this
distillation are known to the industry. However, few distillers actually
have the installed capacity to produce grain neutral spirits; only one in
g
Kentucky has such a capacity. Thus, most distillers would be required to
ship the recovered alcohol to a location with distillation capacity or
install the capacity themselves. Both options present additional costs.
The recovered alcohol would be at approximately 50 proof before
redistillation, and in such a dilute form, would cost 19 cents/proof
gallon to transport by tank truck.10'11 The costs of installing and operating
distillation equipment to produce grain neutral spirits were not
calculated but would be considerable.
4.3.3 OSHA Standards, Insurance, Energy, and Secondary Environmental^ Impact
An important consideration in applying carbon adsorption to whiskey
warehouses is the effect the control device will have on safety and worker
health. Closing the warehouse to install a CA unit could increase the
concentration of ethanol inside the warehouse, potentially violating OSHA standards
and increasing insurance risks.
The OSHA standard for ethanol is 1000 ppm, time-weighted-average for
8 hours. Several of the proposed design cases are based on 1500 ppm ethanol
in the warehouse, an apparent violation of the OSHA standard. However, several
factors should be considered. First, the OSHA standard is a time-weighted
4-11
-------
TABLE 4-2
Distilled Liquor Sales
(10) wine gallons/yr
Vodka
Gin
Cordials
Rum
Bottled Cocktails
Imp. Whiskey
Other
Blended Am. Whiskey
Straight & Bonded
Whiskey
TOTAL
1975
65.0
36.2
101.2
23.8
- 14.4
7.0
95.3
19.4
159.9
46.6
64.1
1973
54.0
35.3
89.3
20.6
13.4
5.0
91.9
17.3
148.2
53,5
66.2
110.7
371.8
119.7
357.2
Industrial Ethanol Use
(1016 gallons 190 proof/yr
1975 210
1976 200
1980 220
Ethanol Market Pattern
Percent
Chemical Manufacture 44
Solvent 46
Export 10
4-12
-------
average with no short term maximum exposure limit. Thus, the OSHA standard would
not be violated if a worker spent only part of his time in the warehouse and the
remaining time outside or in other parts of the distilling complex. Thus,
a 1500 ppm ethanol concentration would not restrict entry. The OSHA standard
may affect labor practices since workers could not remain in the warehouse
all day.
Secondly, as the discussion of whiskey quality indicates, the CA system
would of necessity have to be operated to reproduce existing conditions and
practices. The 1500 ppm design case was chosen to represent ethanol
concentration presently used in aging. Thus, the installation of a CA
system would present no additional problems for worker health compared
to present methods of operation.
Contacts with an insurance company indicated that no additional
12
insurance on the warehouse is required. In addition, as discussed
above, the operation of a CA system should not increase ethanol levels
in the warehouse over existing levels.
Another important consideration in control device evaluation is energy
and secondary environmental impact. In recovering ethanol and converting it
to a usable product, the m£.in areas of energy consumption are the steam used
in regeneration of the carbon and in redistilling. Assuming that a one still
system can adequately purify the recovered alcohol, the energy usage for
regeneration is calculated to be 6.6 x 10 joules/kg ethanol recovered and for
redistillation 7.9 x 10 joules/kg ethanol recovered. The energy for redistillation
would be required even without the control system since the recovered alcohol
would be'replacing alcohol presently produced. By comparison, a distiller
in his normal production operations (cooking grain, heating warehouses,
c
operating other stills) uses an estimated 80 x 10U joules/kg ethanol
recovered. In addition, the energy value of the ethylene required in production
of synthetic ethanol is calculated to be 33 x 10 joules/kg ethanol. Thus,
the proposed control system could potentially save energy.
4-13
-------
The main secondary environmental impact of the control system is the
disposal of the waste water from distilling the recovered alcohol to grain
neutral spirits. The amount of waste water produced in this manner would
be 4 liters/kg ethanol recovered. By comparison, using a figure of 143 liters
water/bushel, grain in producing whiskey and assuming 95 of these liters
become waste water, an estimated 61 liters waste water/kg ethanol recovered
is produced by the normal operation of a distiller, Existing methods of waste
water disposal at distillers should be able to handle this extra load.
4.4 CARBON ADSORPTION - WAREHOUSE TESTS
Between 1960 and 1968, a major distiller operated a carbon adsorption
system on a whiskey warehouse at one of their facilities. A second
distiller, National Distillers and Chemical Corporation, also installed a carbon
adsorption system in the early 1950's to develop background data for a patent.
However, the National test was conducted on only one warehouse floor, for one year,
diverting a very small fraction of the exhaust air through a laboratory size
carbon adsorber. Thus, the only full-scale test of the proposed control
system is the one run from 1960 to 1968.
Table 4-3 lists the important data from the full scale test. Several points
should be noted. First, the recovery efficiency and the proof of the
recovered alcohol are both lower than the values used in the design calculations.
Second, the carbon adsorber increased the rates of evaporation from the barrel and
adversely affected quality. This last effect, the alteration of whiskey quality,
was one of the principal reasons the test was stopped.
The full scale test, c.s run, does not demonstrate that a carbon
absorption unit can be successfully applied to whiskey warehousing. At a
recovery proof of 30, the transportation cost for the recovered alcohol is
4-14
-------
Table 4-3. CARBON ADSORPTION SYSTEM DATA
FULL SCALE TEST, 1960-1968
Adsorber Design & Operating Parameters
Warehouse Size/Type:
Barrel Emission Rate:
Recovery Efficiency:
Recovery Proof:
97,500 Barrels/Brick & Concrete
5.25 Ib/barrel-yr
74 percent (5 yr. average)
30.5
Operating Procedures & Conditions
Experiment One (1960-1964)
Ventilation Rate
Recirculation
Humidity
Proof
Whiskey Quality
Experiment Two (1965-1968)
Ventilation Rate: Normal
Recirculation: No
Humidity: Normal
Year 1 & 2
Normal
Yes
Elevated
Decreased
All years
Year 3
Reduced
Yes
Elevated
Decreased
Sour, wet
wood
characted
Year 4 & 5
Normal
No
Normal
Stabilized
Improved to
satisfactory
Proof: Normal
Quality: Poor all years
Chronology: The changes in year 3 of experiment one were made to reduce the
elevated humidity and temperature in the experimental warehouse. This proved
unsuccessful and due to this and continued problems with whiskey quality,
changes were made in year 4. The second experiment was run since the number
of changes that were made in the first experiment made it unreliable as a data
source.
Other Effects:
Evaporation: During both experiments, the rate of evaporation from the barrels
increased. During the first experiment, the increase was .3 percent/yr
(3.2 percent/yr. vs. 2.9 percent/yr normal) and during the second experiment,
the increase was .4 percent/yr higher (3.3 percent/yr vs. 2.9 percent/yr normal)
Recovery: During the first two years of experiment one, when the adsorber
exhaust was recirculated to the warehouse, the recovery rates were 83.3 and
93.3 percent compared to the 74 percent overall recovery for all five years.
4-15
-------
32i£/proof gallon; this amount must be subtracted from the value of the
recovered alcohol since the distiller would be required to absorb this cost.
The recovery rate is 10 percent lower, and the steam usage higher (at 30 proof,
the steam rate is 7 kg/kg) than the figures used in the design calculations,
again adding costs. Finally, the whiskey lost due to the excess evaporation
would need to be reproduced at $2.10/proof gallon aged. Though some of this is
recovered by the carbon adsorption system (75 percent in the full scale test study),
the recovery value is much lower. The effect of these factors on the recovery
system cost is shown in Table 4-4. Thus, the factors in the test result
in a net loss for the system. However, the net loss is 4.Si/proof gallon
aged, compared to $2.10 production costs. Therefore, the increased costs
shown in the test, though significant, do not by themselves make tne system
infeasible.
The more critical problem was the system's demonstrated adverse effect on
whiskey quality. In the full scale test, 360 barrels (180 in the second experiment)
were filled with a quality approved lot of whiskey and split equally between
the experimental warehouse (the warehouse with the CA unit) and a control
warehouse (a warehouse operated normally). Whiskey quality tests were run
yearly on samples from both sets of barrels; the samples were evaluated by
taste test panel in a procedure similar to the method by which the actual
product is tested. The results are shown in Table 4-3. The quality was poor
into year three of experiment one; subsequent changes in the recovery system
corrected this poor quality in year four and five. A second experiment was
conducted to verify these results; however, the quality was poor in all years.
The acceptable quality of years four and five in experiment one seems to have
occurred because the poor quality of the previous years was being "undone."
Normally, aging would not start with whiskey which had an inferior
quality that needed to be corrected.
4-16
-------
Table 4-4. COST CALCULATIONS
FULL SCALE TEST
Design Parameters:
System Parameters:
Costs:
Cost per Proof Gallon
No. of barrels: 100,000
Emission Rate: 5.25 Ibs/barrel-yr
Ethanol Concentration: 1500 ppm (assumed)
Excess loss: .35 percent yr (average of
two experiments) or .35/2.9 =
.12, fractional increase in
emission rate
Recovery: 75 percent
Steam Rate: 7 Ibs steam/1b ethanol recovered
Adsorber size
Adsorber size
Ethanol lost:
Ethanol
Steam:
Carbon:
calculated: 5290 scfm
1.5 x calculated: 7930 scfm
5.88(10)b Ibs/yr
recovered: 4.41(10)^ Ibs/yr,
1.33(10)5 proof gallons/yr
3.09(10)6 Ibs/yr
12,720 Ibs
Annual Capital Cost $20,220
Taxes, Ins., etc. 5,390
Electricity 2,800
Steam 5,250
Maintenance 8,640
Carbon 1 ,820
44,120
Credit for recovered -27^,9JO
ethanol, $.21/pg
(includes transportation)
Cost/final proof gallon
Net cost
Excess Evaporation
Total Cost
55 proof gallons/
barrel orignally
100,000 barrels
minus evaporation
minus soakage
$225(10)3/4.72(10)6
$16,190/yr
$64,760 for 4
years
. 12(100,000)(5.25)4 =
252,000 Ibs, 76,130 proof
gallons at $2.10/proof gallon
$159,980
$224,720 for four years
5,500,000 proof gallons
- 532,000
- 250,000
4,718,000 final proof gallons
= 4.8<£/proof gallon
4-17
-------
It appears that certain changes in the design and operation of the CA system
during the test could have eliminated problems encountered. First,
the low recovery rate experienced was apparently due to the inadequate size
of the adsorber unit. During each cycle, it is hypothesized that the bed
became saturated and breakthrough occurred. Alcohol laden air thus
passed through the adsorber to the atmosphere with no recovery occurring.
The higher recoveries experienced during the first two years were apparently
due to the recycling of the adsorber exhaust stream to the warehouse. Thus,
when breakthrough occurred, the unrecovered alcohol was recirculated back
-into the warehouse and no loss to the atmosphere occurred. This unrecovered
alcohol was eventually captured because, as it was recirculated back to the
warehouse, the ethanol concentration in the warehouse increased. This increased
concentration would increase the capacity of the adsorber unit, resulting in
the eventual recovery of the alcohol. Confirmation of this hypothesis
would require, among other things knowledge, of the adsorber bed capacity at the
concentration, temperature and humidity of the warehouse air. This
information is not available.
The deterioration of whiskey quality in the test study was apparently
caused by three factors: higher humidity, lower ethanol concentrations,
and continuous ventilation. The elevated humidity existed in the first three
years during the time the adsorber exhaust was recirculated. Since the CA
unit did not remove water, the recirculation of the adsorber exhaust resulted
in the accumulation in the warehouse of the water evaporating from the barrels.
The lower ethanol levels resulted from the continuous removal of organics from
the warehouse by the CA unit. Though natural ventilation would also remove
ethanol, the CA unit provided continuous air removal. In contrast, natural
ventilation would be intermittent, removing ethanol only occasionally. In
fact, during nights, weekends and winter, there may be no ventilation in
warehouses since during those periods the windows and doors are sometimes
-------
closed. In addition to continuous ventilation lowering the ethanol
concentration, continuous ventilation also upset the stagnant air layers
that develop around the barrel in natural ventilation. As discussed
in Chapter 2.0, the removal of these stagnant layers replaces the
stop-start diffusion pattern that normally occurs with natural
ventilation.
The manner in which these factors affected quality is not clear. However,
the altered concentrations of ethanol and water around the barrel and the
continuous ventilation probably altered the concentrations, and cycles in
concentrations, of substances in the barrel wood and bulk whiskey. The
rates at which the mechanisms responsible for aging - extraction and solubilizing
of wood constituents, diffusion of these constituents into the bulk liquid,
chemical reactions between the various substances and transport of air into the
bulk liquid - occur depend on these concentrations. Thus altering these
concentrations alters the rate at which the aging mechanisms proceed,
altering whiskey quality.
Various modifications in the test may have alleviated the whiskey
quality problems. These modifications would have been to operate the system
intermittently and to recirculate the adsorber exhaust part of the time.
Intermittent operation could have beer accomplished by sequencing the floors
that receive ventilation, as described in section 4.3.1. Another option would
have been to shut off the CA system during periods when the warehouse windows and
doors would have been closed under normal operation. Such a method of operation
would have allowed for stagnation periods, permitted the accumulation of ethanol
to the proper levels required for aging, and reduced or eliminated excess ethanol
evaporation. Partial recirculation could have eliminated the problem of both low
and excessive humidity. This could have been accomplished by occassionally routing
the adsorber exhaust to the warehouse.- The amount of partial recirculation would
be determined by the humidity level in the warehouse; the adsorber would be
/I-1Q
-------
exhausted outside when the humidity became too high. Another variation of
partial recirculation could occur in winter, when high air circulation
rates may have been required for forced air heating. During this period, the
adsorber could have been partially bypassed, with this by-pass stream being
recirculated. This would allow for sufficient air movement for heating, without
exhausting ethanol laden air to outside and without upsetting aging by
removing the ethanol from the larger air streams required for heating.
4.5 ALTERNATE SYSTEM OF AGING
A novel system of whiskey aging is under development in which maturation takes
place not in charred oak barrels but in closed stainless steel vessels lined with
straight charred staves. This system is of interest due to its potential
for large savings in aging costs and for almost complete elimination of aging
losses. Its applicability to whiskey aging and control of warehousing emissions
will depend on the system's ability to produce whiskey of acceptable quality.
The central component of the system is a cylindrical stainless steel vessel
approximately 5 meters in diameter and 7 meters high, holding approximately 100,000
liters of liquid. Inside the vessel, straight charred oak staves are held in
the whiskey by arms extending radially from a shaft at the center of the vessel=
The staves are arranged so that air spaces created between them are manifolded
together to the central shaft holding the arms, and from there to vacuum, pressure
and condensing equipment. The central shaft can be designed to rotate to move
the staves through the whiskey. The vacuum equipment pulls vapors through the
staves to duplicate aging and the condenser recovers this vapor as liquid
and returns it to the vessel. The pressure equipment provides for further
controls over the aging process potentially useful in producing whiskey
of a desired quality. Finally, internal heating coils provide for temperature
control of the aging whiskey.
4-20
-------
The large cost savings in the system occur in three areas. First,
the labor and wood cost of the barrels is reduced by using straight wood
staves and using less wood per volume of whiskey stored. Second, the loss
of whiskey through evaporation is eliminated since the system captures
the vapors and returns them after condensation. Third, the warehouse
area is reduced since the system requires only 1/1Oth the volume. The cost
savings that result can be substantial, up to 50 percent of present aging costs,
The system's most important feature of the system from an emission
standpoint is the complete elimination of whiskey loss. Loss during
aging is eliminated since ethanol evaporating through the staves is captured
in the air spaces manifolded to the condensers, which return the vapor as
liquid to the vessel. Soakage losses are reduced since the alcohol remaining
in the used staves is partially recovered by continuing to draw a vacuum
after the whiskey is emptied. The vacuum evaporates the ethanol in the
staves and draws it to the condensers where the ethanol is recovered. Finally,
any losses due to spillage and barrel leaks are eliminated since the whiskey
is piped into and out of the aging vessels. Thus, the system has the capacity
to be almost loss free.
The key factor determining the system's applicability to whiskey aging
and emission reduction is the quality of the whiskey produced. Since
testing of the system has not been completed, it is not known if the system
will properly age whiskey. Testing of the system is scheduled for 1978.
4-21
-------
4.6 CONTROL OF BARREL SOAKAGE LOSSES
The major control device discussed to this point, carbon adsorption, is
applicable only to the control of evaporation during barrel storage; control of
losses due to soakage in the barrel staves would require additional measures. These
measures, along with present uncontrolled practices, are described below.
Present practice is to rinse used barrels with one gallon of water before
selling or storing the barrels. The amount of whiskey recovered in this
manner appears to be low since such a rinse removes only the surface
film of whiskey on the barrel staves. One distiller practices a more complete
rinse using 3 gallons of water and rolling and shaking the barrel to improve
recovery. This practice removes approximately one half gallon from the barrel
14
wood, or about .7 kg ethanol. This is less than 20 percent of the estimated
3.8 kg of ethanol in the barrel wood. Thus, present practices recover only
a small percent of the liquid soakage in whiskey barrels. No other systems
to further recover barrel soakage are in practice.
Three types of systems have potential applicability: more complete
rinsing, vacuum evaporation, and steaming. More complete rinsing could be
accomplished using a greater amount of water, greater agitation of the barrel,
more than one rinse and heating the water. Vacuum evaporation would involve
connecting the used barrel to a vacuum source to draw out the vapors. Vacuum is
available at most distillers since vacuum evaporation is used to dry spent
grain for animal feed. Steaming would involve passing steam through the
barrel, using the heat to evaporate the ethanol in the wood. The steam would .;
then be condensed to recover'the ethanol. The dilute whiskey produced in these
methods could be used in adjusting the proof of bottled whiskey. Whiskey is 5
typically diluted before bottling, since it is aged at higher proofs than
those at which it is marketed.
4-22
-------
Two factors appear to limit the effectiveness of all three recovery
methods, the inherent slowness of diffusion in wood and the barrel configuration.
The physical mechanisms, extraction, heat, and vacuum evaporation, on which
the recovery methods are based all attempt to increase the rate of diffusion
of ethanol through the wood. However, the small pore structure of the wood and
the great width of the stave (2 cm is a considerable distance in terms of molecular
diffusion) results in extremely slow diffusion; 3 to 6 months are required to
saturate the wood after filling the barrels. Even if a hundred fold increase in
the diffusion rate could be achieved, more than a day would be required to
recover all ethanol in the barrel staves. In addition, the barrel configuration
does not allow optimum contacting in rinsing and steaming. Water touches only
a small percentage of the wood at any one time in rinsing, and unless extra
holes or special spargers are provided, steam distribution inside a barrel
would be uneven and steam contact with the walls poor.
It would appear that other methods of recovery of barrel soakage losses
might be necessary. These methods would require methods of operation both unfamiliar
to the whiskey industry and complex. They would involve splintering the barrels
into small slivers of wood, passing the slivers through water extraction and
vacuum filtration and evaporation. The slivers would then be available as fuel.
Alternately, the saturated wood slivers or the saturated staves themselves could
be fed to a boiler. Adjustments in the boiler operation would be required to
assure proper firing with saturated wood as a partial fuel. As noted, these
operations would be complex, but could be technically possible and,
with credit for the wood fuel and recovered ethanol, financially feasible.
However, no analysis of this option was made.
One final method may be feasible, storage of the empty barrels in enclosed
warehouses vented to a carbon adsorber. An economic analysis of this option is shown
4-23
-------
on Table 4-5. The analysis assumes that nine months of storage would be
required to remove 85 percent of the liquid in the barrel wood and that
the first 20 percent of the liquid would have been removed by water rinsing.
Thus, assuming 3.8 kilograms of ethanol in the wood, the system would
recover .65(3.8) or 2.5 kg from each barrel. A warehouse ethanol concentration
of 250 ppm was chosen since a low concentration would be required to evaporate
the liquid from the wood. Finally the recovery efficiency was set at
95 percent or better since no special features would be required to protect
whiskey quality. The final cost of the system is 2.8
-------
Table 4-5
Control System for Barrel Soakage
Losses - Warehousing
Assumptions
Recovery on Adsorber
Design
Costs
Storage period:
Ethanol level:
Total Barrel soakage;
Warehouse capacity:
Removal from barrel
95 percent
Emission rate:
Adsorber size:
Surface Area:
Carbon:
Recovery:
Steam:
Annualized Capital Cost:
Taxes, Insurance, etc:
Electricity:
Steam:
Carbon:
Maintenance:
Warehouse-Depreciation
Handling (50<£/barrel )15
Recovery Credit
Net Cost
Cost/proof gallon
15
9 months
250 ppm
3.8 kg ethanol
50,000 barrels
85 percent
20 percent from rinsing
65 percent from storage
3.3 kg ethanol/yr-barrel slot
21,900 scfm
292 ft2
35,040 Ibs
104050 pg
1.03 (10)6 Ibs/yr
$46,000
$12,260
,730
,750
$ 7:
$ 1
$ 5,000
$ 8,640
$15,000
533,330
S129,710/yr
$55,150
$74,560/yr
2.8?
4-25
-------
4.7 REFERENCES
1. U. S. Environmental Protection Agency. Control of Volatile Organic
Emissions from Existing Stationary Sources - Volume I: Control Methods
for Surface Coating Operations. EPA 450/2-76-028, November, 1976.
2. U. S. Environmental Protection Agency. Capital and Operating Costs of
Selected Air Pollution Control Systems. EPA 450/3-76-014, GARD, Inc.
May, 1976.
3. Memo from Ted Jackson, Chemical Manufacturing Section, U.S. EPA, to
David R. Patrick, Chemical Manufacturing Section, U.S. EPA, Research
Triangle Park, N.C. Cost Estimates of Carbon Adsorption. December 19, 1977.
4. Telephone conversation between Mr. Larry Euchre, Union Carbide Corporation,
Cleveland, Ohio and Mr. David Augenstsin, PEDCo, Cincinnati, Ohio,
April 27, 1977.
5. Chemical Marketing Reporter, January 16, 1978, pg. 7.
6. Letter and comments from Mr. Murray Sobolov, Chairman, DISCUS Technical
Committee, to EPA. October 26, 1977.
7. Liquor Handbook. Clark Gavin, editor, Gavin-Jobson, Inc., New York.
1974, 1975, 1976.
8. Chemical Marketing Reporter. Chemical Profile-Ethanol. September 13, 1976.
9. Trip Report from David C. Mascone, U.S. EPA to David R. Patrick, Chemical
Manufacturing Section, U.S. EPA, Research Triangle Park, N.C. on a meeting
with the DISCUS Technical Subcommittee. January 20, 1978.
10. Telephone conversation between Mr. Ted Jackson, U.S. EPA and Mr. Larry Thompson,
Liquid Transporters, Inc., Louisville, Kentucky. November 7, 1977.
11. Telephone conversation between Mr. Ted Jackson, U.S. EPA and Roger Fredrick,
Southern Tank Lines, Langstown, Pennsylvania, November 7, 1977.
4-26
-------
12. Telephone conversation between Mr. Fred Hoit, Factory Mutual Corporation
Norword, Massachusetts, and Mr. Terry Briggs, PEDCo, Cincinnati, Ohio.
May 4, 1977.
13. Murray, Robert G. Preliminary Evaluation of New Method for Aging
Spirits. Standford Research Institute, Menlo Park, California.
14. Trip Report by Terry Briggs, PEDCo, Cincinnati, Ohio, on a visit to
Jim Beam Distillery, Clermont County, Kentucky. April 7, 1977.
15. Reference 13.
4-27
-------
-------
APPENDIX A. EMISSIONS FROM THE PRODUCTION
OF UNAGED WHISKEY
The production of unaged whiskey involves preparation and fermentation of
grain and distillation of the resulting liquid to produce unaged whiskey. The
three largest sources of volatile organic emissions in this operation are the
fermentor vent, the distillation column vents and the drying of the used grain.
The fermentation of grain in whiskey manufacture produces large amounts
of carbon dioxide. This carbon dioxide exits from the fermentor by vents
on the top and carries with it minor amounts of ethanol. A measured value for
3 1 3
this emission is 183 g ethanol/m grain. Using 146 proof gallons whiskey/m grain,
and a production of whiskey of 79.2 x 10 proof gallons in 1976, the total
nationwide emissions from this source are 99 MT/yr. A typical large distillery
producing 4 x 10 proof gallons whiskey/year would emit 5.0 MT/yr.
In the operation of the various distillation columns in a distillery,
ethanol is emitted from the inert vents on the column condensers.
However, with the double condenser system commonly used and condenser temperatures
of 70 to 90°F, these emissions are low. One emission estimate is 0.0022 kg
2
ethanol/proof gallon-column. Using the whiskey production above, and assuming
1.5 columns/distillery as an average, the total nationwide emissions from this
source are 260 MT/yr. A typical large distillery with a 3 distillation column
system producing 4'x 10 proof gallons/year would emit 26.4 MT/yr.
A-l
-------
The grain remaining after fermentation and distillation is typically
dried and sold as animal feed. During drying some of the residual ethanol
in the grain is evaporated to the air. The ethanol content of the grain
slurry remaining after distillation is 0.1 to 0.01 percent by weight;3 however,
a large portion of this ethanol would be mixed with the wastewater removed
from grain slurry. Assuming 0.05 percent ethanol in the grain and that 30 percent
is evaporated to the air, the nationwide emissions are 206 MT/yr. A large
distillery producing 4 x 10 proof gallons/yr would emit 10.1 MT/yr.
The typical large distillery described in this appendix is analagous
to the typical distillery in Chapter 3.0. That distillery had emissions of
1460 MT/yr from aging; the total emissions from the emission points described
in this appendix is 41.3 MT/yr, less than 3 percent of the aging emissions.
A-2
-------
REFERENCES APPENDIX A
1. Telephone conversation between Mr. Lew Heckman, EPA Region II, New York,
New York, and David C. Mascone, U.S. EPA, February 9, 1977.
2. Carter, R. V., and B. Limisky. Gaseous Emissions from Whiskey
Fermentation Units. Atmospheric Environment. 8:57-62. 1974.
3. Telephone conversation between Dr. Allen Thomas, Brown-foreman Distillery,
Louisville, Kentucky and David C. Mascone, U.S. EPA, Research Triangle
Park, N.C., December 22, 1976.
4. Trip Report by Terry Briggs, PEDCo, Cincinnati, Ohio, on a meeting with
the DISCUS technical committee, March 30, 1977.
A-3
-------
-------
APPENDIX B.
WHISKEY BY VARIOUS PERIODS OF PRODUCTION REMAINING IN
BONDED WAREHOUSES IN KENTUCKY AS OF DECEMBER 31, 1975
Prepared from information obtained at the Office of the Department of Revenue of the Commonwealth of Kentucky
DISTILLERY
Barton Brands, Inc.
Bardstown, D.S.P. Ky. 12
JoS. B. Beam Distilling Co.
Bardstown, Kentucky
Beam, Ky.
Clermont, Ky.
Blair Distilling Co.
St. Francis, Ky,
J.T.S. Brown's Son Co.
Lawrencebury, Ky,
Brown-Forman Distillers Corp.
(3 Units) Louisville, Ky.
Commonwealth Distillers, Inc.
(Formerly T.W. Samuels)
Deatsville, Ky.
Double Springs Distilling Co.
Bardstown, Ky.
Frankfort, Ky,
Louisville, Ky.
Flcisuhmann Distilling Corp.
Ovvenshoro. Ky.
Glenmore Distilleries Co.
Owensboro, Ky,
Yellowstone, Inc.
Louisville, Ky,
Heaven Hill Distilleries, Inc.
Bardstown, Ky.
Hoffman Distilling Co.
Lawrenceburg, Ky.
Medley Distilling Co.
Owensboro, Ky.
Ben F. Medley Distillery
Stanley, Ky.
National Distillers & Chem, Corp.
(3 Units) Louisville, Ky.
(3 Units) Frankfort, Kv.
Austin Nichols Distilling
lawrenceburg, Ky.
Jessamine County, Ky.
REMAINING WHISKEY PRODUCED OR RECEIVED
BOTTLED IN BONO - AGE
CALENDAR YEAH ENDING DECEMBER 31
Over
8
Yean
25.829
5,698
12,069
4,450
858
11,299
2,470
1,399
1,243
6,621
13,207
6,768
844
75
1,493
1,411
3,413
1963
No.
Barrel!
10,536
2,122
25,207
24,761
2,783
5,625
8,214
1,642
1,019
203
24.968
3,311
24,058
1,423
1,275
12,258
7,740
16.083
1969
No.
Barrel!
34,533
303
14,981
4,523
23,391
4,321
7,071
4.538
5,928
389
5,412
8,988
10,577
35,726
869
6.759
35
96,993
124.302
23,202
1970
No.
Barred
53,657
1,110
31,594
4,336
10,582
37,320
4,266
7,190
25
35,963
25.111
23,637
49,775
824
3,137
133,920
152,553
20.050
1971
No.
Barrels
34,464
17.572
24,102
328
13,816
60.514
6,540
10,753
30,412
45.418
20.891
66,816
2,099
31,098
119
126.436
151,314
14,635
i
1972
No.
Barrel]
1,544
91,239
84,464
63,371
3,928
16,731
36,411
40,017
18.236
62,141
28,745
99,304
106,923
22,763
1973
No.
Barrels
64,273
98,247
78,559
531
41,340
5,644
15,380
35,413
29.884
13,076
64,771
29,721
23,552
1974
No.
Barrels
16,831
41,233
64,014
74,076
104,437
1,800
38,568
10.816
53,863
17,928
30,225
1975
No.
BarreU
20,248
13,320
58,943
60,743
97,000
30,501
1.117
47,423
9,713
66.605
17.445
16,732
TOTAL
No.
Barrel!
261,931
54,553
339.253
405,795
9,718
82,000
412.444
28,261
38,524
53,633
2,676
213,288
181,007
101,661
417,791
11,933
129,220
223
470,4(M
611.348
171,420
16.732
Per
Cent
4.26
799,601
13.01
.16
1,33
6.70
.46
94,833
1.54
3.47
2.94
1.65
6.30
JO
MO
1
.31
1,03 1.752
17.S3
183,152
3.06
-------
APPENDIX B. (Continued)
WHISKEY BY VARIOUS PERIODS OF PRODUCTION REMAINING IN
BONDED WAREHOUSES i!M KENTUCKY AS OF DECEMBER 31, 1975
Prepared from information obtained at the Office of the Department of Revenue of the Commonwealth of Kentucky---;
DISTILLERY
Old Boone Distillery Co.
Meadowlavvn, Ky.
Old Fitzgerald Distillery, Inc.
Louisville, Ky.
Schenley Industries, Inc.
Bernhsim Distilling Co.
Louisville, Ky.
Park&Ti!fordOist.of Ky.
Louisville, Ky.
TheGeo. T. StaggCo.
Bardstown, Ky.
Frankfort, Ky.
Joseph E, Seagram & Sons, Inc.
Louisville, Ky.
Cynthiarta, KY.
Lawrenceburg, Ky.
Huntington Creek Corp.
Coxs Creek, Ky.
Star Hill Distilling Co.
Loretto, Ky.
Willett Distilling Co.
Bardstown, Ky.
Totals Each Year Dec. 31, 1975
Totals All Years Dee. 31, 1975
Totals Dacember31, 1974
Totals December31, 1973
Totals December 31, 1972
Totals December 31. 1971
Totals December 31,1970
Totals December 31, 1969
Totals December 31,1968
REMAINING WHISKEY PRODUCED OR RECEIVED
BOTTLED IN BOND - AGE
CALENDAR YEAR ENDING DECEMBER 31
Over
8
Yean
14,254
6,107
6,209
6,062
32,634
49,972
12,459
1,752
12,733
462
5,349
247,150
235,498
230,085
177,510
214,333
331,462
413,702
504.299
1968
No.
Barrels
4,783
36,252
27,569
2,679
510
23,492
23,900
3,616
2,575
48,447
1,183
1,271
343,575
Rt!S.9S3
386,813
1,143,734
1.3C6.734
1.428.095
1,196,524
1,731.446
1969
No.
Barrel]
3,725
61,382
38,212
3.922
9,614
31,342
33,558
8,351
1,145
133,235
2,789
4,210
761,557
935.317
1,159,606
1,335,124
1 ,354.3,24
1,462,894
1.653,901
1970
No.
Barrett
1,483
51,119
22,478
14,727
1,284
19.593
L 16.459
4,893
' 3S3
84,539
3,643
5,343
820,990
960.854
1,100,151
1,114,402
1,170,710
1.381,303
1971
No.
Barrels
269
50.417
21,692
2.991
43,242
26,330
2,143
75
53,969
4,334
4,711
863.700
1,018,144
1,014.776
1,070,059
1,171,353
1972
No.
Barrels
2,142
38,420
53,988
5,543
10,428
92,417
17,593
661
40,305
fi.OOl
75
343,395
943,578
1.024.001
1.081,542
1973
No.
Barrel*
9.812
10,369
103,108
9,767
18,222
114,147
5,303
1,389
25,791
6,491
2,875
313,766
846,142
1,004,877
1974
No.
Barrels
3,314
3,962
44,387
16,135
10,309
58,934
11,089
34,424
5,637
3,942
657,580
743,722
197S
No.
Barrels
3.997
9,287
47.43S
- 19,713
133,601
21,825
4,975
4,522
685,564
TOTAL
No.
Barred
43,780
273,915
370,673
58,335
105,711
587,240
174,576
22,820
4,164
439,443
36.125
37.328
6,148,587
6.683,654
7,285,998
7,514,642
7,877,969
8,491,893
8,609,815
3,706,688
Per
Cent
.71
4.45
1,102,515
17.93
641,003
10.43
.59
.61
Note - Fractional barrels reduced to one full barrel. Storage does not necessarily represent ownership.
-------
TECHNICAL REPORT DATA
("lease read Interne ni^'is on Me rci-ersc before Completing)
NO.
EPA-450/2-78-013
3. RECIPIENT'S ACCESSION NO.
M. TITLE AND SUBTITLE
Cost and Engineering Study - Control of Volatile
Organic Emissions from Whiskey Warehousing
|5. REPORT DATE
April 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHOFUS)
David C. Mascone, ESED
a. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION ftlAME AN.D. AODHESS
U.S. Environmental Protection Agency
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park,
North Carolina 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This document provides cost and engineering information on control of
volatile organic emissions from whiskey warehousing. Included are a description
of whiskey aging, warehousing, and of volatile organic emissions from warehousing;
a development of emission factors and inventories for these emissions; a cost
and engineering analysis of available control techniques for these emissions;
and a discussion of volatile organic emissions from other whiskey manufacturing
operations. The major control technique discussed is carbon adsorption.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Air Pollution
Whiskey Warehousing
Whiskey Aging
Ethanol
Carbon Adsorption
Air Pollution Control
Stationary Sources
Organic Vapors
8. DISTRIBUTION STATEMENT
Unlimited
j 19, SECURITY CLASS (This Report) 21 . NO. OF PAGES
! Unclassified
20. SECURITY CLASS (This Dug
unclassified
EPA Fo.-m 2220-1 (Rev. -J-77)
-------
INSTRUCTIONS
1. REPORT NU.vlBER
'risen ;ne L.PA report number as it appears OQ thi; cover of ihe public-ision.
2. LEAVE BLANK
3. RECIPIENTS ACCESSION NUMBER
Reserved for use by each report recipient.
A. TITLE AND SUBTITLE
Title should indicate clearly and briefly the subject coverage of the report, and be displayed prominently. Set subtitle, it used, in smaller
type or otherwise subordinate it to main title. When a repor! is prepared in more than one volume, repeat the primary title, add volume
number and include subtitle for the specific title,
5. REPCRT DATE
Each report shall carry a date indicating at least month and year. Indicate the basis on which it was selected (e.g., date of issue, date of
approval, date of preparation, etc.).
6. PERFORMING ORGANIZATION CODE
Leave blank.
7. AUTHOH(S)
Give name(s) in conventional order (John R. Doe, J. Robert Doe, etc.). List author's affiliation if it differs from the performing organi-
zation,
8. PERFORMING ORGANIZATION REPORT NUMBER
Insert if performing organization wishes to assign this number.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Give name, street, city, state, and ZIP code. List no more than two levels of an organizational hirearchy.
10. PROGRAM ELEMENT NUMBER
Use the program element number under which the report was prepared. Subordinate numbers may be included in parentheses.
11. CONTRACT/GRANT NUMBER
Insert contract or grant number under which report was prepared.
12. SPONSORING AGENCY NAME AND ADDRESS
Include ZIP code.
13. TYPE OF REPORT AND PERIOD COVERED
Indicate interim final, etc., and if applicable, dates covered.
14. SPONSORING AGENCY CODE
Insert appropriate code.
15. SUPPLEMENTARY NOTES
Enter information not included elsewhere but useful, such as: Prepared in cooperation wilh, Translation of, Presen'.ed'at conference of,
To be published in. Supersedes, Supplements, etc.
16. ABSTRACT
Include a brief (200 words or less) factual summary of the most significant information contained in the report. If the report contains a
significant bibliography or literature survey, mention it here.
17. KEY WORDS AND DOCUMENT ANALYSIS
(a) DESCRIPTORS - Select from the Thesaurus of Engineering and Scientific Terms the proper authorized terms that identify the major
concept of the research and are sufficiently specific and precise 10 be used as index entries for cataloging.
(b) IDENTIFIERS AND OPEN-ENDED TERMS - Use identifiers for project names, cod« names, equipment designators, etc. Use open-
ended terms written in descriptor form for those subjects for which no descriptor exists.
(c) COSATl I'lELD GROUP - Field and group assignments are to be taken from the 196.5 COSATl Subject Category List. Since the ma-
jority of documents are multidisciplinary in nature, the Primary Field/Group assignment(S) will be specific discipline, area of human
endeavor, or type of physical object. The application(s) will be cross-referenced with secondary Field/Group assignments that will follow
the primary posting(s).
18. DISTRIBUTION STATEMENT
Denote releasability to the public or limitation for.reasons other than security for example "Release Unlimited." Cite any availability to
the public, with address and price.
19. & 20. SECURITY CLASSIFICATION
DO NOT submit classified reports to the National Technical Information service.
21. NUMBER OF PAGES
Insert the total number of pages, including this one and unnumbered pages, but exclude distribution list, if any,
22. PRICE
Insert the price set by the National Technical Information Service or the Government Printing Office, if known.
EPA Foro 2:20-1 !Rev. 4_/7' -:Re
-------
-------
ENVIRONMENTAL PROTECTION AGENCY
General Services Division (MD-28)
Office of Administration
Research Triangle Park, North Carolina 27711
POSTAGE AND FEES PAID
ENVIRONMENTAL PROTECTION AGENCY
EPA-335
OFFICIAL BUSINESS
AN EQUAL OPPORTUNITY EMPLOYER
Return this sheet if you do NOT wish to receive this material I I
or if change of address is needed | ]. (Indicate change, including
ZIP code )
PUBLICATION NO. EPA-450/2-78-013
------- |