-------


 c



















itonng instruments
:e according to the
i 1
E Q,
— 0
< =
eel
specifications.
CJj
c
c
c
c
(N

-------
4. Recordkeeping and reporting requirements

       Tables  2. 2 and 2. 4 identify the recordkeeping and reporting requirements that apply-
when using the CAPE+RTO System as an alternative means of emission limitation to meet the
requirements of 40 CFR Part 63, Subpart II.
                                         2-14

-------












*~*
(vi
0)
.a
o
CAPE+RTO Syst
i.788 identified in
«> "
O (o
£ c
i- O
o +-
"- 0
CO Q)
«• C/7

0) C
li
3 O


a .2
en '5
c cr
H
r>^ to
Lt o
TO =
re Q.
.ES
a- c
'able 2.4 Recordkee
(in additioi
r—











60
C
'•E
o
Q.
OJ
a:
3
C
C
re
CJ
















Recordkeeping

















tu
re
E
C/5
C
o
E
tB
3
u-
O
'to
c.
o
c

^_
0
^
c
'c
o
o
)»-
o
c"
•.s
0
c/:
C.
CJ
c
1
TO
60
_O
Jo
C
ll
11 o.
TO • =
3 "£;
"So
IB C
£ O
TO '-p
Q S
o.
— re




u

c
^
IB
tB
~
fc_
TO
CJ
e enclosur
m taken.
•£ •-
"4- t3
O re
3 "O
3  c
o -a a
c; c IB
re c
CJ ™ 0
- oo re
13 £ £
'TO c E
T3 re ,>
IB '-^ TO
CJ TO -Q
H O tn


G
CJ
re
T3
C
re

I
0
•o
1»
G
O
O
a:
^
c
o
CJ
o
G
* >
4. Values of the air flow ratf
each measurement.

IB
E.I
£ ^j
^* re . f
C? T3 ^
< TO fc
5 £ o
0 § fe.
w ao
•o £ .E

.2 f w
2 '5 o
*- O" •*-
C CJ O
ij i
Ofl ^ W5
ral ^ t±
t- _ c:
o>! £ i>
i- .£ |
a: E -E
£ SJ .s
*" o n
| « §
-§| | g
T3 ^ -^
5e . II
g ^ 5 1 1
c cj _a: TO •r;
H •£ S D re





_rz
c u
• ™ *+*
C cj
II
' T3
»^ re
^3 CJ
O "^"
•u ~°
C 73
TO ^
entrations at RTO inl
id connection to RTO
0 C
C TO
5. Values of the VOHAP co
assembly of CAPE sections
measurement.


u'
f c
'^
2

u
J=
^u
b.

^
fc
o
T3
CJ
E
3
«
TO
^ C
0^
H 2
flj O
J= «
*- o
II
CB re
ej ^
"S g
H ^


_c

're "^
t: re
C- cj
< 13 .
!_> >-. oo
• * i .E
u 'i S
C*. O t?
60 " _ra
^™ ^'
"*lj ^ ^—
z^z
° -^ G
_c t— c:
•fc G G
O ^
uj "— S u
t. ^v IB ^3
^J fli ** re
_s| E tu cj
•5 -s e c
,s: "o ra ^
"^ C t« /*>
^ o^ ™ ^^
^^ ? —
6. Operating Time/Air teml
Record the date, start time (1
temperature reading within 1
extra time (t2) needed to run



•o

.—
cr
£
0
~
"
(O c

ll
E ra
2 "°
ti §
? C
0 |
cj • -
•£ «-"
TO +
S cj
•1 •-
T^3 bj)
c c
re ~
| 1
f- c












PM
+_
RTO operating time.
+
u
a.
U
T3
.£
'5
o
"re












S
2
CJ
a.
o
O
Bi
O
'ra
•O
T3
C
TO
O
.£
T3
C
OJ
•o
C
re
C
TO
55














tes of RTO operation.
TO
T3
•u
C
TO
IB
CJ
.E
-o
•o
c
TO
TO
55
00


















Z


U
cj (y
•£ ^
iS01
--g
S TO
3 t>
"So
0 IB
G cj
Q. re
"T3 L^
C
w .„;
§1
11
r's QA/QC recommen
le to Table 3 of this a
I- TO
9. A copy of the manufactu
instruments and units applic
check.

(N

-------







00
c
C
o
CL
O
Q£
"re
3
c
c
re
'i
oc
























CO
'E.
 — —
u re
o c
S..2
u re
u j-
c H
re g-
c 0
o 4>
c "
1 1
T1 "1^
c £
c 2
O X
11
U C^
CL D.
o B-
 0
























<;

Z

c
o
•^
re
c
o
E
u
o
-o
to
_c
c
c.
CL
3
tA
CJj
c
•5
_2
u
c;
"re
|
CL
CL
re
l_
o
c:
o
'i
_o
d-
CL
re
o
f.

v«
o
-^.^
&
o
u
<

-
o
L_
u
CL
re
CL
_C
•a
o
tJ
'i
XI
3
c/>
O
»^
wLJ
fe1
E
(/)
t
o
CL
£
O
c/j
"0 «J
° S
<-> P
"1
U v-
S .H
re c
0 O
'"" k-
CL o
CL 
< "S

X
« wi
« re
•5 u
10 5.
tn ^
•0 >0
u ^^
O tfi
o re
-------
                                  Section 2.2
                             Element  C(1), Parti
 "The owner or operator of an affected source may apply to the Administrator for
permission to use an alternative means such as (an add-on control system) of limiting
emissions from coating operations."

Section: 63.783(c) also specifies that an application for alternative means of limiting emissions must
include:

(i) "An engineering material balance evaluation that provides a comparison of the emissions that
would be achieved using the alternative means to those that would result from using coatings that
comply with the limits in Table 2 of this subpart, or the results from an emission test that
accurately measure the capture efficiency and control device efficiency achieved by the control
system and the composition of the associated coatings so that the emissions comparison can be
made..."

      When you use both capture and destruction unit operations you need to perform an emission
test.  The test should define the following three values:

      (1) The capture efficiency of the enclosure.
      (2) The destruction efficiency of the add-on control unit operation.
      (3) The amount of time (in hours) the destruction unit needs to be operating after each coating
cycle. This parameter is especially important for air-cured (dried) coatings and when "time averaging"
of emissions from coatings is not permitted.  You cannot use time averaging of emissions when it results
in  exceedence of an individual coating limit on a solids (nonvolatiles) basis.

2.2.A  Capture Efficiency of CAPE (enclosure)

      MMC contracted with Pacific  Environmental Services (the Contractor) to perform air emission
tests on a ship, the USS Scott DDG-995, during the period  August 19-23,  1996 at MMC shipyards in
Norfolk Virginia. A copy of the 1996 Emissions Test [2] report was submitted together with the 1996
Application for Approval [3] and the 1996 Implementation Plan [4].  The MMC followed, as will be
shown later, an operational protocol that would satisfy all but one of the requirements for 100 percent
capture efficiency.
                                          2-17
-------
       MMC also investigated the effectiveness of the CAPE (enclosure) for controlling emissions
associated with abrasive blasting of ships in dry dock and the results were discussed in the 1996
Emission Test report. MMC identified, in the report, other environmental benefits that may be achieved
by using the CAPE (enclosure) during ship repair operations.

2.2.B   Destruction Efficiency of the RTO

       MMC meets all requirements for destruction efficiency based on detailed and well documented
performance test data involving gaseous organic compounds, using EPA Method 25A [5].  The 1996
Air Emission Test results show that RTO can achieve:

 •      Destruction efficiencies greater than 98 percent, when the concentration of VOC pollutants is
       greater than 100 ppmv.

 •      Efficiency of 90 percent when the concentration is 50 ppmv (propane equivalent).

 •      Output concentration from the RTO much below the 50 ppmv value, which represents the cut-
       off value for using EPA Method 25A.

       In Table 5 of the 1996 Application for Approval MMC says that they will operate the RTO to
produce a destruction efficiency of 95 percent, determined using EPA Method 25A, the level they will
want to take credit for in future operation. We have explained the conditions under which you can use
EPA Method 25A in a  1995 memo [5].

2.2.C   Operating Time for the RTO

       MMC tried to develop a simple procedure to relate the time it would need to run the RTO when
they apply coatings with different assumed solids (nonvolatiles) content. In their calculations, they

•      Assumed that coatings released the volatiles instantaneously.

•      Average the emissions from coating over time, concluding
       that they  could turn off the RTO while still applying coatings.

However, we have concerns about the reliability of MMC's predicted RTO operating values because
under certain conditions the assumptions may not be valid.  For example, they did not consider the
effect of temperature or season on evolution of volatiles. Considering all these factors, we identified
the appropriate operating time  for the RTO in Table 1.1 above.

       Also, an operator must not turn off the air to the RTO or the  RTO itself when applying a
coating. We cannot assume when you stop applying a coating that all the necessary mass of volatiles
was swept out of the enclosure and directed to the thermal oxidizer.  Under certain application
temperatures and coating VOHAP content, you will need to operate the RTO beyond  the time it takes


                                            2-18
-------
to complete the application of a coating (Chapter 3). The necessary time is defined at the point when
the total emissions inside the CAPE System divided by the total solids (nonvolatiles) on the hull is at
the level or below that of each complying coating. If we accepted the argument that the higher
destruction efficiency achieved during certain painting cycles should balance any time-period of
noncompliance resulting from the application of coatings in the enclosure—we would inadvertently be
granting those that have elected to use add-on controls in lieu of using complying coatings a
compliance-related advantage.  The requirements are given in Section 2.1.

2.2.D  Coating Limits Never-to-be-Exceeded Form of the Standard

       The NESHAP specifies under Section 63.783(c)(iii)(2) that the post-control emissions of
VOHAP per volume of applied solids (nonvolatiles) must be no greater than those from the use of
coatings that comply with the limits in Table 2 of Subpart II (Table 1.2),  meaning that the post-control
emissions at any time should not exceed the mass of VOHAP per volume of nonvolatiles (solids)
applied.  Since the limitations in  Table 2 of the standard are specified on a never-to-be-exceeded
basis, the use of non-complying coatings is not permitted and constitutes a violation of the standard,
unless the emissions released are being controlled to the level of the compliant coatings—referenced on
a solids or nonvolatiles basis.

     When we developed the regulation in 1993, we had evaluated several  approaches for determining
maximum achievable control technology (MACT) options. All were based on coatings with
inherently low pollutant contents [6]. Two options were evaluated during the development of the rule:

       (1) The first type of limit is based on maximum or never-to-be-exceeded values for each
coating category. The facility and paint manufacturer would know that by using or producing a coating
that, as applied, meets the MACT limit, he or she is complying with the regulation. You are not
allowed to apply in uncontrolled environment, coatings that emit VOHAPs above these limits.

       (2) The second type of limit is based on averaging. Average limits  allow the shipyard to use
any coating,  but they must do extensive planning, calculating, and recordkeeping to make certain they
meet the average limit. Use of any high-HAP coating must be offset by use of one with low-HAP
content within the designated averaging period.

       Of the options evaluated for selecting the "floor," the never-to-be-exceeded basis was the
option adopted in the final regulation based on the comments received from the stakeholders [7-9].
As a result, certain existing coatings which are non-complying cannot be used, once the regulation was
in place, without first installing add-on controls.  The never-to-be-exceeded limits were the basis for
calculating the MACT floor and the emission reduction achievable resulting from this regulatory
action.  Any  form of time-averaging that resulted in exceeding a limit is therefore not permitted.

       Time-averaging of emissions would, thus, allow a source to use low or zero VOHAP/VOC
coatings during certain painting cycles and non-compliant coatings during other coating cycles. You
would be in compliance,  as long as the total emissions released during the averaging period did not

                                            2-19
-------
exceed the total emissions produced had only complying coatings been used throughout that time
period. However, this is not permitted. When you are using the CAPE+RTO System you are allowed
to use non-complying coatings and the emissions from coating cycles may mix inside the CAPE . You
will need to operate until the mass of VOHAPs within the CAPE divided by the solids applied to the
hull is less or equal to that of the complying coating.

2.2.E VOC as Surrogate for VOHAP

      MMC indicated in its application that it has selected the option allowed in the NESHAP of
using VOC content as a surrogate for VOHAP content.  The term VOHAP content as used in this
standard includes any volatile emitted during air curing of the coating. However, any 'exempt'
compound that is an organic HAP (non-VOC HAP) will need to be counted when determining the
total mass of VOC material that would be emitted from a coating.
                                          2-20
-------
                                  Section  2.3

                              Element C(1), Part 2
 (ii)  "A proposed monitoring protocol that includes operating parameter values to be monitored
for compliance and an explanation of how the operating parameter values will be established
through a performance test..."

       In this section we will evaluate the 1996 Emissions Test (performance test) [2] results for the
CAPE+RTO System. In the first part we discuss the extent MMC satisfied the 100 percent capture
efficiency requirements defined in EPA Method 204 (62 FR 32500, June 16,  1997) [10]. In the second
part we evaluate the procedure followed by MMC for setting the value of the destruction efficiency of
the RTO.

2.3.A CAPE (enclosure)

       The Unit Operation System (UOS). the ensemble on which the material balance should be set to
determine the change in the VOHAP concentration resulting from application of a coating, is the CAPE
(enclosure) volume plus the air volume in pipes of the air management system (Figure 2.3).  If the
CAPE system captures in-time the necessary amount of VOHAPs and passes these VOHAPs to the
RTO, the CAPE system will be considered to have 100 percent capture efficiency.

       The requirements of a Permanent Total Enclosure and for ensuring  that the  flow is into the
enclosure are specified in EPA Method 204.  If the following four requirements (Requirements 1 to 4)
are met and if all the exhaust gases from the enclosure are ducted to the add-on control device
(Requirement 5), then the VOHAP capture efficiency is assumed  to be 100 percent, and capture
efficiency need not be measured.  If part of the VOHAP laden gas stream is not ducted  to the add-on
control device, capture efficiency must be determined.  The following paragraphs present each
requirement (total five) and discusses whether the CAPE met or was not able to meet the requirement
during the 1996 Emission Test in Norfolk, Virginia.

       (1)  "Any [natural draft opening] NDO shall be at least four equivalent opening diameters
from each VOHAP emitting point unless otherwise specified by the Administrator."

        MMC indicated on page 2-14 of the 1996 Emission Test report [2] that CAPE design contains
no windows. It has no NDOs such as those that allow raw materials to enter and products to leave and
the doors were normally closed during coating operation (Appendix, Exhibit 2.1).
                                           2-21
-------
       An  NDO according to Section 3.1 of EPA Method 204 is "[Ajny permanent opening in the
enclosure that remains open during operation of the facility and is not connected to a duct in which a
fan is installed."

       MMC determined the equivalent diameter on each side of the CAPE to be  5.2 cm (0.17 ft).
This value comes from worst case assumption that a gap 2.54 cm by 16.5 m (1 in. by 54 ft) existed
between the CAPE and the ship hull, on each side of the curtain touching the ship hull
(Appendix, Exhibit 2.2).  Four equivalent diameters would be 21 cm (0.68 ft).  This means that the
CAPE would meet requirement 1. Under operating conditions the nozzle of a paint gun would
normally be positioned at a distance greater than 21  cm from either  sides of the canvas edges touching
the hull, to avoid painting the walls of the enclosure.

       To evaluate the Capture efficiency of the CAPE, MMC estimated the areas through which the
air infiltrates into the CAPE.  They referred to these areas as NDOs. These areas (NDOs) are located
mainly along the sides of the enclosure where the curtains come in contact with the hull of a ship.
Although the contour of a ship makes it difficult to obtain a good seal, a visual inspection during the
test performed by MMC indicated no visible cracks or openings at the seam of the total enclosure and
the hull.

       (2)  "The total area of all NDOs shall not exceed 5 percent of the surface area of the
enclosure 's walls, floor, and ceiling."

       MMC established the enclosure area ratio (NEAR) for the CAPE at 0.005 using Equation 2.2
below. This value  is below the limit of 0.05 set in EPA Method 204. They arrive at this value by
estimating the following:

       (a) AN. the NDO total area as 23.9 m2 (257 ft2), (Appendix,  Exhibit 2.2), assuming that a
2.54 cm (1 in.) gap existed along ever}' seam of the enclosure.

       (b)  AT, the total enclosure surface area as 5082 m2 (54, 825 ft2).  Five percent of this area is
100m2 (2, 741 ft2).  Hence.

                           NEAR= (AN)/(AT)                                 (Equation 2.2)
                                 =  (23.9)7(5082)
                                 =  .005 (or 0.5%)

 Therefore, the enclosure meets requirement.

       (3)  "The average facial velocity (FV) of air through all NDOs shall be at least 3,600 m/hr
(200 ft/min). The direction of air flow through all NDOs shall be into the enclosure." A vacuum
(negative pressure  drop) of 0.013 mm Hg (0.007 in. H20) corresponds to an FV of 3,600  m/hr (200
ft/min). (NOTE:   0.13 mm Hg (.007 H20) vacuum  is the value in the latest version of EPA
Method 204.)
                                            2-22
-------
Figure 2.3 CAPE and Air Management System Unit Operation System
           (boundary for material balance)
                                                                 312m3/hr
              Make up
              Air
CAPE
6375 m3
                                                 1388m3/hr
                                    1700m3/hr
System Boundary

      The requirement here can be satisfied in two ways:  (1) if the average FV for a permanent
enclosure is at least 3,600 m/hr or alternatively, or  (2) the vacuum inside the enclosure is greater than
0.013 mm Hg. (.007 in. H,0).

      MMC satisfied the alternative option (2). It continuously monitored the pressure drop (static
pressure) along four centrally located positions in the CAPE (Appendix, Exhibit 2.3). Each of the
probe locations was: 9.15 m (30 ft) high; 1.53 m (5 ft) from the outer wall of the enclosure; and  18.3 m
(60 ft) between either  air plenum. The positions were equally spaced, 18.3 m (60 ft) from one another.
Location number 1 is about 73.2  m (240 ft) from the exhaust point, and location number 4 was the
closest at 18.3 m (60 ft). The static pressure was always above the required minimum of-0.013  mm Hg
(-0.007 in H20) throughout the four days of the monitoring period: The average static pressure for the
four positions along the CAPE was -0 .035 mm Hg (-0.019 in. H20). Requirements is, therefore,
met.
                                         2-23
-------
       In the following section we will show you the results we obtained when we tried to use the
equation in Method 204 shown below to calculate FV:

                      FV = ( Q0 -Q,) /AN                                       (Equation 2.3)

where:

       Q0 = the sum of the volumetric flow from gas streams exiting the enclosure through the exhaust
       duct or hood, corrected to standard conditions. [ Q0 = 1700 nrYmin ]

       Q, = the sum of the volumetric flow, corrected to standard conditions, from all gas streams into
       the enclosure through a forced makeup air duct; zero, if there is no forced makeup air into the
       enclosure. [Q; = 1388 mVmin]

       AN = total area of all NDOs in enclosure. [AN = 23.9 m2]

       The value of FV for the CAPE is equal to [60 (1700 - 1388)] / 23.9 = 780 m/hr (43 ft/min).
This value is lower than the 3600 m/hr (200 ft/min) minimum value specified in EPA Method 204.
Since MMC. in its 1996 Emissions Test,  indicated that when the seal between the CAPE and hull was
inspected visually, there were no visible cracks or openings, one can conclude the worst case
assumption that a one-inch gap existed at every seam of the total enclosure which overstimated the total
area An of all NDOs.  MMC did not perform this calculation as it did not select this option. It met
instead the negative pressure drop requirement.  We expect the air flow through the CAPE to be
inwards when we operate the enclosure at this minimum vacuum. [10]

       (4)  "All access doors whose areas are not included in Criterion Number 2 and are not
included in the calculation in Criterion Number 3 shall be closed during routine  operation of the
process."

       This criteria is met. In this set up, there are two doors positioned, one at each end of the CAPE.
 The>T are used to access or depart the enclosure. The criterion is met because the doors are normally
closed. There are no windows to the enclosure.

       (5)  "All VOC emissions must be captured and contained for discharge through a control
device. "

       This criterion was not met. MMC is able to discharge using this setup all emissions from an as-
applied coatings that would need to be captured and sent to the oxidizer.  However, MMC's procedure
presented in the 1996 Emission Test (performance test) and which it indicated in its Application for
Approval would not do that. We explain in Section 4.2 why the approach used by MMC for
determining RTO operating time would not guarantee that the necessary amount  of VOC emissions are
captured and contained for discharge.
                                            2-24
-------
 2.3.B Regenerative Thermal Oxidizer (RTO)

       If we use an add-on control unit such as an oxidizer, the owner or operator needs to do an initial
performance test to show that the required emission reduction is achieved. Another purpose of the test
is to identify and validate the monitoring protocol that includes operating parameter values to be
monitored for compliance.  MMC provided detailed results of a test undertaken in Norfolk, Virginia, in
1996 [2]. They used several test methods:.

•      EPA Method 1 of Appendix A of 40 CFR Part 60 was used for sample and velocity traverses.
•      EPA Method 2 of Appendix A of CFR Part 60 was used for velocity and volumetric flow rates.
•      EPA Method 3 of Appendix A of CFR Part 60 was used for gas analysis.
•      EPA Method 4 of Appendix A of CFR Part 60 was used for stack gas moisture.
•      MMC proposed using EPA Method 25A to measure VOC concentration in the inlet duct to the
       RTO. and in the outlet of the RTO.  It conducted three 1-hour test runs using one flame
       ionization (total) analyzer (FIA) at the inlet, and one FIA  at the outlet of the RTO.  It also
       indicated that during a test, the coating application rates reflected near maximum operating
       condition.
•      MMC determined the destruction efficiency of the incinerator. Equation 2.4 below identifies the
       parameters that were measured in the demonstration test.

               DE = (C, Q, -C0Q0)/ C,Q,                                         (Equation 2.4)

where:

       DE = destruction efficiency, %
       C, = inlet VOHAP concentration, ppmv propane, dry basis
       C0 = outlet VOHAP concentration from incinerator, ppmv
           Propane, dry basis
       Q, = inlet  flow rate. nvVminute at STP
       Q0 = outlet flow rate, m3/minute at STP

       MMC conducted a system bias check prior to conducting  the coating test runs to verify that
each entire sampling system was leak tight.  It also conducted a calibration error test within 2 hours of
initiating a test run. Zero and span drift checks were performed at the completion of each sampling run.
These tests are not required in EPA Method 25A. It shows that MMC was paying attention to quality
control.

       On page 13 of the Application for Approval [3] MMC indicated that it will use the approach in
the  1996 Emission Test to demonstrate compliance with the NESHAP using the CAPE+RTO System
(Appendix, Exhibit 2.4).  However, EPA will allow MMC to use the 1996 Emission Test [2] in lieu of a
new performance test if they are using the same CAPE+RTO System in the new locations and operating
as explained in Section 2.1.
                                           2-25
-------
(iii) "Details of appropriate recordkeeping and reporting procedures".

       The facility indicated that it will maintain separate records for the coatings applied to a substrate
within the CAPE system, including the amount of each coating, the volatile content (including cure
volatiles), and volume solids (nonvolatiles) of the coating applied to the ship hull. This will not be
required when the CAPE+RTO System is used as indicated in Section 2.1.

       When a coating is thinned, the thinning allowance should be calculated to meet the limits on a
solids basis.  That should be the basis for the as-supplied and the as-applied certifications. This is
explained in Section 63.785 (c)(2) & (c)(3) of the regulation. MMC recognized that Equation 1 in the
standard [1]  needs to be used to determine thinning allowances.  However, they are checking
compliance using g/liter of coating units in the certification sheet for "noncomplying coatings" (page 19
of their 1996 Implementation Plan). They should have used g/1 solids  limits instead. Their statements
in Section V(c)(2), page 8 of the Application for Approval indicates that compliance be based on "g/1 of
coating or g/1 of solids." They need to indicate that g/L solids (nonvolatiles) shall be used whenever
thinning solvents are added to coatings.  MMC indicated that it will make the indicated revisions.

       Lastly, MMC put much thought into planning the work and reporting the test information in the
1996 emission test performed in Norfolk, Virginia.  They documented well how they  measured
emissions of VOC material and particulates. The test report contains supporting details, including
essential quality control logs and data tables (e.g.. Appendix, Exhibits 2.5  and 2.6).   However, I found
it difficult at times to match pieces of data together.  Such issues were  later clarified (Appendix,
Exhibit 2.7).
                                             2-26
-------
                                 Section 2.3

                                  Element C(2)
"The Administrator shall approve the alternative means of limiting emissions if, in
the Administrator's judgment, postcontrol emissions ofVOHAPper volume applied
solids will be no greater than those from the use of coatings that comply with the
limits in Table 2 of this subpart."

       We will discuss some of the material balance related information submitted by MMC. The
purpose is also to clarify a number of important points that have application beyond the MMC's
submittal.

2.4.A  Material Balance Calculations  of Theoretical Minimum VOHAP Reduction 76

       MMC used a material balance to calculate, for a representative number of coatings applied to
ships, the minimum reduction of emissions, referenced on a solids (nonvolatiles) basis, that is needed to
meet the coating limitations in the regulations. They did the calculations for two coatings with assumed
solids (nonvolatiles) content from 20 to 80  percent by volume.  The results are shown in
Table 5 of their 1996 Application for Approval [2], (Appendix. Exhibit 2.8). MMC concluded, for
example, that if the calculated reduction necessary was 77 percent for a given  coating mixture, using an
RTO with a destruction efficiency of 95 percent would achieve a higher reduction than is required by
the regulation. MMC referred to this minimum reduction as an "overall control efficiency required."

EPA Comments

We disagree with their conclusions.

•       The calculated minimum reduction  of VOHAPs on a nonvolatiles basis (solids) is a useful
       number.  It does not however, indicate the overall control efficiency achieved at a given RTO
       setting, unless we can assume that

            "all the paint emissions evolve at once and all the air flow directed
            from the enclosure (CAPE) into the RTO occurs instantaneously."

•       The above scenario is unlikely and MMC will need to factor in the amount of time needed for
       the necessary amount of volatiles to reach the RTO from the enclosure.

                                         2-27
-------
2.4.B  Material Balance as an Alternative to Emission Testing (Model 1)

       MMC used another material balance approach (Model 1) to estimate the VOC build-up within
the CAPE during application of a coating. They calculated the hours of RTO operating time required
for compliance for coating application periods varying between 2 and 8 hours (Appendix, Exhibit 2.9).
The calculations were performed for two types of coatings, in nonvolatiles (solids) content from 20 - 80
percent and presented in a table, which was to be used as an alternative to emission testing.  They also
calculated another table to illustrate the time needed to achieve concentrations of less than 908 g
(2000 Ibs) (Appendix, Exhibit 2.10).

1.  EPA comments (Model 1)

       We believe that MMC started with a good idea, however, the model they used does not
represent the situation under study. Our reasons are explained below:

•       An implicit assumption in the model used is that all VOHAPs (or VOCs) are flashed off
       instantaneously when the paint is applied. This would lead to a higher rate of evolution of
       VOHAP material. It would also result in predicting a shorter time for reaching the maximum
       VOHAP concentration than might occur in an actual coating cycle.

•       The Company did not restrict the use of Table 2 in the 1996  Application for Approval
       (Appendix, Exhibit 2.9) to one temperature and one would conclude that the table was going to
       be used throughout the year.  A coating will take much more time to dry in winter than in
       summer.

•       Table 2 in Exhibit 2.9 shows that for antifoulant coatings with 40 percent solids (nonvolatiles)
       by volume, applied for 8 hours, the  RTO would need to operate for only 5 hours  to bring down
       the emissions to the level of compliant coatings. This implies that MMC can turn off the RTO
       while they are still applying coatings; however, we disagree as explained in Section 2.2.

•       MMC assumed that the NESHAP permitted the time-averaging of  emissions to meet the
       limitations in Table 2 of this NESHAP.  This explains why they indicated that the RTO could be
       turned off before they completed painting and still be in compliance. The NESHAP  does not
       allow averaging of emissions. In Table 2 of the regulation (Subpart II) the limits are specified as
       never-to-be-exceeded.

•       Although MMC did not provide details about the model, it was possible to duplicate some of
       the results by assuming, as a starting point in the analysis,  that the mass  of volatiles removed and
       the mass of volatiles recirculated were proportional to the flow.  That assumption did not lead
       to the conclusion reached by EPA using a more robust model (Model 2). Hence, MMC should
       not use Table 2 in Exhibit 2.9 of the Appendix to meet the standard.
                                            2-28
-------
2. EPA comment (reduction to background level. Appendix. Exhibit 2.10)
   (Table 4 of the 1996 Application for Approval)

•      An owner or operator of the RTO is only required to ensure that the "postcontrol VOHAP
       emissions" per volume of applied solids be no greater than those from the use of coatings that
       comply with the limits  in Table 2 of the NESHAP (Table 2.1) and not the destruction of all the
       applied VOHAPs.

•      An owner or operator of the RTO is not required to reduce emissions to background level.
       Therefore, MMC's did not need to develop their Table 4 (Appendix, Exhibit 2.10). The table,
       however, provides a reference point for the maximum achievable reduction-background level.

 •     MMC showed that the RTO would need to operate 14 hours-6 additional hours after painting
       ceases to achieve VOHAP background level. At this point the coating should be at a minimum
       dry-to-touch, but  it will still retain a small percentage of the solvent. This can range from
       5-15 percent by mass volatiles (Appendix, Exhibit  2.11). It may take between 4 to 36 hours
       from the time a coating is applied to be dry. i.e., dry-to-touch (near zero emissions). The drying
       times depend on temperature and type of coating and other operating parameters. The terms
       dry-to-touch and dry-to-hard are explained in some detail in Chapter 3.0.

2.4.C  Determining Operating Time for the RTO

       The values in Tables 2 and 4 of the Application presented by MMC underestimate the operating
time for the RTO for reasons explained in Chapter 3.0  and, therefore, should not be used. MMC must
modify the CAPE supervisor log, the CAPE+RTO System operator log. and the CAPE air compressor
operating log. to include the total planned hours of coatings, which ends with each volume of non
complying coating applied.

       The RTO operating time for compliance =  t, + t2. hours are indicated in Equation 2.1, where,
t, = coating cycle time (hours) plus t, = additional time to ensure that RTO is not stopped before the
emissions in the CAPE from the coating cycle are equal to or below that of a complying coating.

       Ideally, an owner or operator should determine the value of t2 at the recommended dry film
thickness and  temperature as explained in Chapter 3.0.  MMC is already collecting this type of
information as part of the Quality Assurance Record for Critical Coated Areas (Appendix, Exhibit 2.5)
for the batch of coating.  However, MMC will be required to follow a simple approach for determining
t2 It will first need to measure temperature inside the CAPE and use this value to determine the extra
time from Table 1.1. The operator will have to choose between one of two locations:
       Inside the air conduit exiting the CAPE.
       At some height in the middle of the CAPE.
 The specific requirements are explained in Section 2.1 Element C(3).

                                           2-29
-------
REFERENCES

1.  40 CFR Part 63, Subpart II--National Emission Standards for Shipbuilding and Ship Repair (Surface
Coating).

2.  Air Emission Evaluation Total Gaseous Organic Compounds and Filterable Particulate Emissions
Compliant All Position Enclosure (CAPE) System USS SCOTT DDG-995 Metro Machine
Corporation; prepared by Pacific Environmental Services, Inc., Herndon, VA, September 1996 for
Metro Machine Corporation, Norfolk, VA.

3.  Application for Approval of Alternative Methodology for Compliance -with The NESHAPfor
Shipbuilding and Ship Repair; submitted by Metro Machine Corporation, Norfolk, VA, June 12,  1996
(Revised October 31, 1996); prepared with Pacific Environmental Services, Inc., Herndon, VA.

4.  Implementation Plan for Compliance -with the NESHAPfor Shipbuilding and Ship Repair Metro
Machine Corporation; prepared by Eric Lasalle, November 1, 1996; Metro Machine Corporation.
Norfolk, VA.

5 .  Memo from John B.  Rasnick, April 4, 1995, "EPA 's VOC Test Methods 25 and 25A,"
EMC GD-033, URL: http://www.epa.gov/emc.

6.   NESHAP for Shipbuilding and Ship Repair; Docket No. A-92-11; Documents No. II-C-014.

7.  NESHAP for Shipbuilding and Ship Repair; Docket No. A-92-11; Document No. II-D-065.

8.  NESHAP for Shipbuilding and Ship Repair; Docket No. A-92-11; Documents No. II-D-067.

9.  NESHAP for Shipbuilding and Ship Repair; Docket No. A-92-11; Documents No. II-D-069.

10. Preparation, Adoption, and Submittal of State Implementation Plans: Appendix M, Test Methods,
204 A-204 F; 62 FR 32500 (Monday. June 16, 1997).
                                          2-30
-------
                                 CHAPTER 3
         EVOLUTION OF VOLATILES DURING PAINTING
      This chapter will look at typical phases that a coating undergoes after it is applied to a substrate
and that produce emissions and explains why MMC's methods of calculating emission levels must
change to meet our conditions for approval.

3.1 Application and Drying Phases of a Coating

      Protective coatings are usually applied as a system, which may include a primer, a middle coat,
and a topcoat.  Coatings must reach the right conditions before another coat goes on; otherwise, the
new layer will affect the rate of solvent evolution resulting in problems such as blisters.  The coating
manufacturer provides guidance to avoid such problems.

      When a coating is applied to a substrate, a part of the volatile material flashes off during
application or spraying. We do not know how much evaporates in the first phase because it depends on
the operation. The rest of the volatile material evaporates in two distinct drying phases. The first is the
"dry-to-touch"  or "dry-to-tack" time, which can vary from 2 hours to much more than 10 hours. The
second is the "dry-to- hard" time, which may vary from 4 to 72 hours.  These relatively long drying
(curing) times occur when a drying oven is not used or the air is much below oven level temperatures .
(Oven temperatures are usually greater than 90°C or 194°F). A paint needs more time to dry
completely (dry hard), but we do not need to know the time referred to as the dry-to-hard time,
because little solvent emits during that period.

Dry-to-Touch Time

      We can define the dry-to-touch time (drying time), as the time until the paint is still soft when
touched but does not stick to one's finger. A Material Safety Data Sheet (MSDS) will state the "dry-
to-touch" time as the recommended film thickness. As the ambient temperature decreases, drying time
markedly increases (Appendix, Exhibit 3.1).  This drying time is important because it marks the point at
which the coating on the substrate contains no more than 3-20 percent by mass volatile material
(Appendix, Exhibit 2.11).  More forgiving paints can receive a second coat at that point.  Usually,
operators recoat such coatings while the paint on the substrate is much more moist (contains more
solvent) than at the dry-to-touch point.  That is why the EPA's model focuses on the dry-to-touch time.
During operations, a coating's drying time depends on a number of factors, including the resin  used in
the coating, the thickness of applied coating, the substrate's temperature, and the air flow.
                                           3-1
-------
3.2  How Long Volatiles Take to Emit from an Enclosure

       The regulation on shipbuilding and ship repair identifies 23 categories of (marine) coatings that
serve the different applications and functional needs of the industry. They include alkyds, inorganic
zincs, and epoxies. Painters apply coatings to large surface areas by spraying, brushing, or using rollers.
When they paint a ship's hull, they normally leave the coating to dry in the open.  Ducting the emissions
from the hull of a ship to an abatement unit operation has not been practical until recently.

       The CAPE+RTO System (Appendix, Exhibit 3.2) includes an enclosure that captures and
releases volatiles from coatings applied in cycles usually lasting from 1 to several hours. Appendix,
Figure 2.3 of Section 2.3 (shows a schematic of an enclosure with recirculation. A certain mass of air
leaves the top of the enclosure so the system pulls in the same amount of make-up air (fresh air)
through the cracks between the ship's hull and the CAPE's walls. When painting begins, the make-up
air mixes with the volatile material and the recirculated air. The volume of make-up air pulled in is
likely to be affected by the amount of volatiles evolved.  Air should  recirculate fast enough to allow
rapid mixing of the VOHAPs evolving within the CAPE system.  Otherwise, we cannot assume the
concentration inside the CAPE and that leaving in the flow to the RTO are the same.

       The solvent loss from a coating film depends on the temperature, air velocity and turbulence
over the surface, ratio of surface to volume, and other factors that determine rate of evaporation from a
coating film [1].  For most of the coatings discussed here evaporation controls the emissions of
volatiles. After a film forms on the coating there comes a point where the rate of solvent loss from the
coating film becomes dependent on how quickly the solvent can reach the surface of the film to
evaporate. The amount evaporated is determined by diffusion of volatiles between particles in the film
(the diffusion-control regime) [2].  Sometimes, people refer to this period as the dry stage because the
coating film on a substrate feels dry-to-touch.  We are not interested in that second phase.

       Figure 2.1 of Section 2.1 shows the concentration profile (build up and decay curves) for VOC
material during several coating cycles in a coating period that lasted for several days.  In some cases the
curves overlap,  showing that several coating application cycles were occurring at the  same time.  It also
includes intervals when paint was not applied.  For example, a large time gap between the first two
coating cycles exceeded the time-to-dry for the first coating.

3.3  Determining the  Concentration of Volatiles Inside a Coating Enclosure

       In the following  sections we will describe two ways to predict the concentration of volatiles
inside the enclosure while applying a coating:  MMC's method and  our method. Once we know the
likely concentration, we can determine how long the regenerative thermal oxidizer (RTO) needs to run
in order to comply with the limit 571 g/L (4.76 Ib/gal) for a general use coating.  Both methods  assume
that there are no concentration gradients inside the enclosure and that volatiles do not re-enter the
enclosure. As discussed later, we can keep volatiles from leaking out by maintaining a minimum
vacuum inside the CAPE. For this discussion we assume that all volatiles in the coatings are VOC
material.

                                             3-2
-------
 1. Calculation by MMC (Model 1)

       Exhibit 3.3 in the Appendix summarizes MMC's results.  On the same page you will find
variables for a maximum application rate of 72,640 g/hr (160 Ib/hr). MMC determined they had to run
the RTO for several hours beyond the coating cycle time to reach compliance.  (They did not commit to
run the RTO beyond that long.)

        They did not describe the procedure exactly, but we were able to duplicate the initial
 15-minute intervals. To do so, we assumed the mass of pollutants resulting during each interval split
according to the same ratio as the flow going to the incinerator and the flow returned to the CAPE.
MMC assumed that volatiles (VOC) material evaporated instantaneously once sprayed. Also, they did
not factor in their calculations how temperature or season affects volatiles' rates of evolution from a
coating.  Ideally, to determine the volatiles concentration or change in VOC concentration within the
enclosure (CAPE), MMC should have used the total air volume in the CAPE plus piping.  They did not
report this value, but they told us the inside volume of the piping from the incinerator to the CAPE was
less than 425 m3 (15, 000 ft3), which, in this case, is negligible compared to the CAPE volume. We
credit MMC for having considered the need to run the RTO beyond the coating cycle, but their
approach cannot be used to predict the effect of temperature or coating thickness on the volatiles rate
of evolution from an applied coating.

2. Model used by EPA to calculate rate of evaporation (Model 2)

       We calculated emissions using a spreadsheet (Model 2) that was developed by EPA [3].
Model 2  is a combination of two published models.  The first provides exact solutions to the indoor
concentration during and after the application of a coating [1].  The second describes a method to
estimate the source decay rate constant which is calculated on the basis that 90-percent of the solvent
mass has been emitted  at the end of the drying time [4].  We mentioned some of the limiting conditions
and explain other details in the cited paper. Table 3.1 gives the variables that we must define to
determine the volatiles rate of build-up inside the CAPE. They include the air-exchange rate (exhaust-
flow rate and booth volume), an important parameter that affects the volatile's maximum peak
concentration. We will assume in the following example that all volatiles are VOC material. We
determine the application rate and hull area painted from the Appendix, Exhibit 3.4,  as MMC
recommended in a recent communication [5].  We also assumed the dry coating thickness was  127
micron (5 mil) to determine volume solids (nonvolatiles) as was indicated in the Appendix, Exhibit 3.4.

       We used the conditions in Table 3.1 to generate the tabular results in Table 3.2. We can deduce
from these results that, after one hour, the VOC in the CAPE's air is 3.9 g/m3 and the VOC on the
hull's  surface is 51,674 g. The total VOC inside the booth is 76,280 g (75 percent of content of the as-
applied coating).  At this point, the total VOC per liter of solids on the hull is 683 g/L solids (Table
3.2). If we add the amount that is still in the booth air, we get 1009 g/L solids (on the hull).

       If the coating were an antifoulant, the complying limit is 765 g/L solids. The ratio of total VOC
inside the CAPE divided by solids (nonvolatiles) on the hull would reach that limit about 1.7 hours into

                                             3-3
-------
the coating cycle of 2 hours.  Hence, if workers suddenly removed the enclosure, the emission rate
would never exceed 765 g/ L solids on the ship's hull.
        However, a general-use coating with a complying limit of 571 g/L solids would need close to
2.3 hours or about 20 minutes above the time for completing the coating cycle represented here.

Table 3.1  Input Parameters Used in Model 2
Parameters
CAPE * (enclosure) Volume
Exhaust flow rate
Surface Area
Volatiles density
Nonvolatiles (solids) content
Paint 90 % drying time
Spraying Period
Spraying (application) rate**
Overspray
Values
6,371 cubic meters
21,420 cubic meters/hour
1,208 square meter
840 g/L
42.2 % V/V
2 hours
2 hours
256 L/h
30 %
225,000 cubic feet
755,700 cubic feet/hour
13,000 square feet
7.0 Ib/gal.
42.2 %V/V
2 hours
2 hours
67.5 gal/hr
30%
* Neglecting 425 m3 (15,000 ft3) of piping will slightly underestimate VOC evolution time.
** Rounded to three significant figures.

       In Table 3.1 we used 2 hours instead of 2.75 hours, the value recommended by MMC [5].  A
value of 2 hours agrees better with their 1996 test results plotted in Figure 2.1 in Chapter 2.  MMC did
not clearly specify the nonvolatiles content of the as-applied coating. We derived the value in three
steps as follows:

•      First, we calculated an overall coverage based on the coating that was applied inside the CAPE
       (including the overspray):  1209 m2/511 liter = 2.37 nrVliter.  The values we used are from
       Appendix, Exhibit 3.5.

•      Second, we adjusted the value determined in Step 1 above for 30 percent overspray:
       2.377 (1-0.3) = 3.38  mVliter.  This value is the practical (effective) coverage of the ship hull.

•      Third, we used the practical coverage and the dry film thickness (DFT) in the equation below to
       determine the volume fraction nonvolatiles (NV):
                           (1000/DFT) *  NV  =3.38
(Equation 3.1)
                                              3-4
-------
       The DFT was assumed to be 1.27 * 10 '4 meters (5 mils).  If we had used the total sprayed
coating (as-applied basis) to calculate coverage (2.37 mMiter) we would have underestimated the
content of nonvolatiles in the as-applied coating.  However, if the actual overspray was less than the
default value of 30 percent we used in Table 3.1, we would be underestimating the NV fraction of the
coating. A decrease in the value of effective or practical coverage (for a given DFT) results in a lower
value of NV content (fraction) and a higher as-applied VOC content which results in a higher VOC
spray rate.

Table 3.2 Volatile to Solids Ratio, g/L (for a 2 hour coating application period)
VOC {g}
Solids (L)
Time (h)
0
0.04
0.08
0.12
0.16
0.2
0.24
0.28
0.32
0.36
0.4
0.44
0.48
0.52
0.56
0.6
0.64
0.68
0.72
0.76
0.8
0.84
0.88
0.92
0.96
1
1.04
1.08
On Surface
On Surface
Ratio 1

1124.4
1099.1
1074.6
1050.8
1027.7
1005.3
983.5
962.4
941.9
922.0
902.7
883.9
865.7
848.0
830.8
814.1
797.9
782.1
766.8
751.9
737.4
723.3
709.6
696.3
683.3
670.7
658.4
in Booth
On Surface
Ratio 2

1610.7
1578.5
1546.9
1515.9
1485.6
1456.0
1426.9
1398.6
1370.9
1343.8
1317.4
1291.6
1266.4
1241.8
1217.9
1194.5
1171.7
1149.5
1127.8
1106.6
1086.0
1066.0
1046.4
1027.3
1008.7
990.6
972.9
                                            3-5
-------
1.12
1.16
1.2
1.24
1.28
1.32
1.36
1.4
1.44
1.48
1.52
1.56
1.6
1.64
1.68
1.72
1.76
1.8
1.84
1.88
1.92
1.96
2
2.02
2.04
2.06
2.08
2.1
2.12
2.14
2.16
646.5
634.9
623.6
612.6
601.9
591.4
581.3
571.4
561.7
552.4
543.2
534.3
525.6
517.1
508.9
500.8
493.0
485.3
477.8
470.5
463.4
456.5
449.7
439.5
429.5
419.7
410.1
400.8
391.7
382.8
374.0
955.7
938.9
922.5
906.5
891.0
875.8
861.0
846.6
832.5
818.8
805.4
792.4
779.7
767.3
755.1
743.3
731.8
720.5
709.5
698.8
688.3
678.1
668.1
653.6
639.3
625.4
611.7
598.3
585.1
572.2
559.6
NOTE: Limits for "general use" coatings is 571 g/L solids, tl +t2 = 2.14 hours
        Limits for "antifoulant" coatings is 765 g/L solids, tl +t2 = 1.65 hours

Required time for CAPE+RTO System is 2 hours. No extra operating hours are required
as shown in Table 1.1.

       Table 3.3 below provides some interesting results calculated by our model:  The wet film
thickness (Ds in Table 3.3) based on total coating sprayed in the enclosure, is equal to
4.24 * 10 "* meters. When we correct this value for 30-percent overspray, we get the practical or actual
WFT, which is equal to 2.48 * 10 ^ meters (~ 10 mil). The Vst in Table 3.3 is the rate of surface (m2)

                                             3-6
-------
coverage and Cw is the VOC content per cubic meter of coating multiplied by (1 minus percent
overspray).  The product Vst, Ds, and Cw is the effective VOC application (spray) rate—grams of VOC
per hour that land on the hull per hour.

Table 3.3  Output Parameters from Model 2 (Reference 3)
            (input conditions in Table 3.1)
 Effective paint spray rate
 Effective VOC spray rate
   179 L/h (excluding overspray}
  87005 g/h
 Paint overspray rate
 VOC overspray rate
 VOC overspray rate

 No. of data points (1)
 Time interval (1)
 No. of data points (2)
 Time interval (2)
   76.8 L/h
44.39C4 L/h
  37288 g/h

  50.00
   0.04 h
   100
   0.02
N =     3.36 (air exchange rate, 1/h)
k =     1.151 (1st-order decay rate const., 1/h)
 kl    0.00049 (1st-order constant, m/h)
 k2      3.36 (air exchange rate, 1/h)
 Ds  4.24E-004 (film thickness, m)
Vst       604 (application rate, mz/h)
 Cw   339864 (VOC cone, in coating, g/m3)
 Calculation of wet film thickness
 paint volume                512 L
 paint volume              0.4238 L/m2
 paint volume           4.24E-OO4 nrrVm2
                        B
                 Vst Ds Cw
        Z21
     87005.18
3.4  Effect of Assuming that the Volatiles Flash-Off Instantaneously

If we assumed that all the volatiles/VOCs are flashed off instantaneously at application, they would
evolve at a higher rate, and we would predict a shorter time to reach the maximum concentration than
would occur in an actual coating cycle.  To illustrate the point, compare the two curves shown in
Figures 3.1 and  3.2, which we produced using our model (Model 2) [3]. The first curve
(30 percent overspray) in each figure shows the time it takes for the concentration to peak inside the
CAPE, assuming 30 percent of all volatiles evaporate instantaneously (flash-off) while the coating is
being applied. The second curve shows the time to peak concentration assuming volatiles evaporate
instantaneously. The first curve better represents an actual situation. When the overspray is assumed
to be 100 percent we can simulate the situation where all solvent flashes off at application.
                                               3-7
-------
Figure 3.1  Effect of Percent Evaporation on Volatile Emissions Profile
            (input parameters are defined in Table 3.1, Temperature ~ 32 °C or 90 °F)
                       A: VOC Concentration in the Booth
                                Overspray 30 %
                             1         2
                                 Elapsed Time (h)
                         B: VOC Concentration in the Booth
                                Overspray 100 %
                                  Elapsed Time (h)
      We compare in Table 3.4 the values of VOC in the enclosure at two critical RTO operating
times: 1.7 hours and 2.3 hours, respectively, for antifoulant and general use coatings. We have also
included in Table 3.4 the values of the VOC concentration in the enclosure .
                                            3-8
-------
Table 3.4 Mass of VOC Remaining in Enclosure
           (Model 2: input conditions in Table 3.1)
Critical RTO
Operating Time
(hours)
1.7 hr
(antifoulant coating)
2.3 hr
(general use coating )
Mass of VOC Remaining in Enclosure
(VOC Concentration in Enclosure Air)
30 % overspray
96, 3 10 g
(4.9 g/m3)
72,356 g
(3.8 g/m3)
100 % overspray
17,43 11 g
(5.8 g/m3)
13,511 g
(2.1 g/m3)
Comments
1. The mass of VOC
remaining when we
assume 100 % flash-off is
underestimated by a factor
of more than 5.
2. The mass of VOC in
the enclosure is not directly
related to concentration in
the enclosure atmosphere.
The VOC content on the
painted surface changes
with overspray level.
       Model 2 calculates the concentration build-up of volatile material by taking into account the
emission change as painting progresses per unit area of substrate.  We assume the coating applies to the
surface at a constant rate. One can visualize that our model divides the application area into many
squares which begin to emit solvent as they are being covered with paint according to an infinitesimal or
"microscopic" emission rate.  The model then combines the effects on these squares to produce a
"macroscopic" emission rate.  This approach explains why our model above shows the build-up of
volatile (VOC or VOHAP) material inside the CAPE taking a longer time to reach the peak
concentration value: it adds the emissions from layers of drying coating to those from the newly
applied coating. (The decay curves are similar for both models.) Of course, we are assuming that the
room air in the enclosure does not cause feedback or back-pressure effect that would lower the
microscopic emission rate as the concentration increases in the enclosure. In reality, such back-pressure
effects may occur. Both Model 1 and Model 2 base their calculations on a coating that will dry to
touch in 2 hours near around 32°C  (90°F) and at that point it will still retainlO-percent by mass
volatiles. However, only Model (2) enables us to readily investigate the effect of ambient temperature
and other parameters on the rate of evolution of volatiles in a coating in their 1996 Emission Test
(performance test) shown in Figure 2.1. The effect of temperature is  reflected in the value of kl, the
first-order decay constant, 1/h. In our model we define kl in terms of time-to-dry (td) in hours [3]:
                                   kl=-ln(0.1)/td
(Equation 3.2)
                                             3-9
-------
3.5 General Discussion

       The model we used (Model 2) to predict the rate of emissions of volatiles from the application
of coatings predicts the volatiles profile for Intergard Epoxy Red that matches reasonably well the
emission data MMC recorded during their 1996 Performance Test.  This provides a level of validation
of our model. The experimental curve based on MMC's data is shown as the first curve on the left in
Figure 2.1 and the predicted curve is shown in Figure 3.1(A).  Model 2 is useful as it makes it possible
for us to predict the effect of operating variables such as temperature and amount of coating sprayed on
the VOC remaining inside the enclosure or on the painted surface.  Figure 3.2 provides a time profile
for the VOC on the coated surface (lower curve) and total VOC inside the enclosure (top curve).  Total
VOC includes the amount on the hull surface plus what is the enclosure air.

Figure 3.2  Effect of Percent Evaporation on Volatile Emissions/Solids on Hull Profile
            (input parameters are defined in Table 3.1, Temperature ~ 32 °C or 90 °F)
                             VOC / Solids on Hull
                                   30 % Overspray
                      O
                      O
                                                       Total VOC
                                                        VOC on hull
                                 0.5   1   1.5   2  2.5
                                 Elapsed Time (h)
       Lastly, most of our discussion dealt with conditions similar to the MMC's test conditions
(Table 3.1) for Intergard Red Epoxy, shown as the first curve in Figure 2.1. However when we apply
(spray) a coating at colder temperatures the drying time in Table 3.1 will need to be increased. This
means that an operator will need to operate the RTO for a longer time to meet the standard as we
indicated in Chapter 1. The total amount of coating applied in a coating cycle will also influence the
amount of VOC remaining in the enclosure.  In its application MMC had based its calculation on what
they referred to as pertinent worst case data: maximum application rate of 72, 640 g/hr (60 Ib/hr),
Appendix A, Exhibit 3.6. This application rate is lower than in the simulation which used some of the
actual data provided by MMC. We obtained a value 90, 218 g/h when we increased the value of the
application period to 2.75 hrs, the value they asked us to use.
                                            3-10
-------
REFERENCES

1. H. Zeh, K. Kohljammer, and M. Krell; VOC-emission from Latex Paints and Plasters During
Application, Surface Coatings International, Vol 4, 1994 pp 142-151.

2.  Z.W. Wicks, F.N. Jones, and S.P. Pappas; Film Formation, In Organic Coatings: Science and
Technology; (Vol. 1: Film Formation, Components, and Appearance); John Wiley and Sons, Inc., New
York, 1992 pp 35-48.

3. The spreadsheet was developed for this project by Dr. Z. Guo, U.S. EPA/NRMRL/APPED/IEMB
(MD-54), Research Triangle Park, NC 27711.  The last version (September 22, 1998) was used.

4. W.C. Evans, Development of continuous-application source terms and analytical solutions for one-
and two-compartment systems, In Characterizing Sources of Indoor Pollution and Related Sink
Effects, ASTM STP 1287, Bruce A. Tichenor, Ed., American Society for Testing and Materials, pp.
279-293.

5. Communication from J. Berry, Environmental, to M. Serageldin, EPA/ESD, October 23, 1998, in
response to questions regarding CAPE+RTO System, Fax from EPA dated October 19, 1998.  This
document will be added to Docket A-92-11.
                                          3-11
-------
                                  Chapter 4
                         Background Information
4.1 Chronology of Events Leading to Approval

      The shipbuilding and ship repair NESHAP (40 CFR 63, Subpart II) was promulgated on
December 15, 1995 (60 FR 64330) and existing sources are required to be in compliance by
December 16, 1997 (61  FR 30814). The NESHAP requires either use of coatings which do not exceed
specified volatile organic HAP (VOHAP) limits, or use of an alternative means of reducing emissions.
Any alternative means must be approved by the Administrator and approval is to be based on criteria
listed in  Section 63.783 (c) of the NESHAP and any applicable requirement specified in Table 1 of
Subpart II.

      On June 12, 1996, MMC submitted to EPA Region III an Application for Approval to use an
emission capture and destruction system in lieu of using compliant coatings to meet the requirements of
the NESHAP. The 1996 Application for Approval, dated June 12, 1996, was first  amended in October
1996 [1]. The submittal included an Implementation Plan dated November 1, 1996 [2] and an emission
test performed in Summer 1996 [3]. A consultant  was contracted by MMC in April 1997 to help bring
the Application for Approval  in line with the standard. Since April 1997, EPA has reviewed several
corrected pages submitted by the consultant. An initial approval plan was defined requiring a
performance test at every site the Compliant All Position Enclosure (CAPE) plus air management
system and the Regenerative Thermal Oxidizer (RTO) unit operations (CAPE+RTO System) were to be
used. MMC requested in March 1998 that it be granted a waiver from having to do a performance test
at ever}'  site considering the cost burden. The elements for such an approach were worked out between
MMC and EPA for several months. During that time MMC requested to revise the operating
parameters in its original submittal, which was based on the 1996 Emission Test (performance test)
results. The parameters changed included the flow to the thermal oxidizer and the minimum vacuum to
be maintained in the enclosure. Consequently, MMC will not be required to meet the exact conditions
of the 1996 Emission Test.

4.2 How the CAPE System and Thermal Oxidizer (CAPE+RTO System) Operate 2

      The control system consists of two unit operations: an enclosure (CAPE) unit with the
air-management system (CAPE) and a regenerative thermal oxidizer (RTO). The CAPE capture unit
consists of several modular sections (tower assemblies) linked to make up the enclosure which
          The documents Metro Machine Corporation (MMC) submitted contain the information.

                                          4-1
-------
conforms to sections of a ship's hull (Appendix, Exhibit 4.1) [3]. In its demonstration test, MMC used
15 tower assemblies to form a hull enclosure (CAPE) around the bow of the ship in Norfolk, Virginia,
encompassing a volume of 6,372 m3 (225,000 ft3). They need four set-ups to complete a ship of the
size used in their demonstration test in 1996 on a military ship, dry-docked in MMC's shipyard location
in Virginia.  Workers must assemble and disassemble a CAPE (enclosure) several times before the hull
area of a ship is painted or before transporting it to a new location (Figure 2.1).

       The following description comes from MMC's 1996 Application for Approval and other
material they have submitted [1-4]. The description helps explain some of the operating variables we
have required MMC to monitor and record.

       MMC's 1996 Application explains that a barge floating next to the ship that will be worked on,
contains the incinerator and the air management system used to clean and circulate the air to the
enclosure (Appendix A, Exhibit 4.1). The barge provides mobility to on-site locations and flexibility to
adapt to various types and sizes of drydocks. It contains equipment to circulate, filter, dehumidify, and
heat the air.  Before MMC starts a coating operation (one or more coating cycles), it starts two blowers
on the barge, combined they draw from the CAPE enclosure 1700 mVmin (60, 000 ftVmin)  of air.  The
air passes through a dust collector that can filter out particles nearly 0.5 microns in size. From here
1400 nrVmin (49,000 ftYmin) recirculate back to the CAPE enclosure, whereas the remainder of the air
flow vents into the atmosphere at 312 m3/min (11,000 ftVmin). A volume of fresh air equal to the
vented air leaks through the cracks between the enclosure and the hull or cracks between the modular
sections (Figure 2.3 and Appendix, Exhibit 4.1).

        During application of coatings and curing, the vented air laden with VOHAP material passes
through the RTO. During the coating operation, the vented air is directed through the oxidizer and
heats to 788°C (1450°F).  A volume of fresh air equal to the air directed to  the RTO leaks through the
cracks between the enclosure and the hull, assuming the enclosure has no significant tears. The
example discussed here represents  a possible operating condition. The actual operating conditions used
during the 1996 Emission Test are  summarized in the Appendix, Exhibit 4.2 (Table  2.3 in Ref. 3.)  The
RTO destruction efficiency under the operating condition in the 1996 test was around 98 percent, when
the pollution concentration was greater than 100 ppmv (measured as propane). The  air in the CAPE
(enclosure) recycles about four times in an hour (volume of air circulated through CAPE/volume of
CAPE). The air turnover will be more frequent in smaller enclosures, if the volume of circulated air
remains around 1700 mVmin (the level during the 1996 Emission Test). This value  is unlikely to change
significantly unless MMC uses blowers having a capacity/rating or a CAPE volume with a verv ditferent
volume.

       MMC controls the modular enclosure's temperature and humidity by monitoring and regulating
the conditions of the supply and return air stream from the equipment on the barge.  During  hot
weather, the "Kathabar" dehumidifier controls humidity and helps lower the temperature. During cold
weather, a hot water heater raises the air stream's temperature. Because the air supply is on a floating
barge, MMC will use special ducts to compensate for tidal and wave motions. For example, slip joints
                                             4-2
-------
and gimbals will keep ducts connecting the equipment barge and a floating or dry dock from breaking
when the barge moves.

4.3  Benefits of Using the CAPE

        No one tested this technology as a unit until 1996, so we did not recommend this technology
as an option to control emissions of particulates and VOHAP or VOC materials while applying
coatings when we developed the regulation on shipbuilding and ship repair or the document that
discussed Alternative Control Techniques (EPA 453/R-94-03).

We see three main benefits. Using the CAPE+RTO System can:

•      Reduce VOC and VOHAP materials resulting while applying coatings.
•      Reduce particulate matter resulting from metal cleaning to remove old coatings from ship hulls.
•      Provides a sheltered environment in which MMC can independently control temperature and
       humidity.

4.4  Design  Features

       An enclosure design determines the ventilation rate which in turn establishes the size of the
heating, cooling, and other equipment, as well as the enclosure's operating cost usually expressed as
dollars per volume of ventilation air [5].

       An  enclosure must contain the largest work piece and provide work space for the workers.
(That explains the CAPE size.)  A recirculation design requires the quantity of incoming fresh air and
exhausted air to be the same.  The fresh air must dilute the air in the enclosure below the 25 percent of
lower explosion limit (LEL).  National Fire Protection Association 33 requires this level to avoid
explosions. The amount of coating applied during a coating period (one or more coating cycles)
determines  the rate of air removal and thus the enclosure's size [5].

       In the  1996 Application  [2] MMC said that  they would limit the rate of coating application to
72,640 g/hr (160 Ib/hr),  so they  would  need to remove 312 nrVmin (11. 000 fWmin) of polluted air
from the CAPE and  direct it to the RTO to destroy the pollutants. The rate of air removed from the
CAPE defines the rate of fresh air that infiltrates the CAPE through cracks and other such openings.

(When MMC applies a coating it reduces the air infiltration because of  the emissions generated during
the coating  cycle.)

       We  have agreed  not to hold MMC to their intended rate, however, because they subsequently
asked to vary the exhaust flow from the CAPE between 284 to 397 standard mVmin (10,000 to
14,000 standard ftVmin). These values bracket the  average actual flow rates used in the 1996 Emission
Test (Appendix, Exhibit 4.2). We will require the facility to follow the  equipment manufacturer's
recommendations for the air flow to the RTO and for operations of the RTO.

                                            4-3
-------
4.5 Construction Features (observations during EPA's August 1998 site visit)

       We could not see the gaps between each of the towers and between the ship's hull and the
enclosure.

       We had to open a large door at the CAPE's entrance to enter it. The exit was a flap at the other
end of the CAPE. They marked it "emergency exit only" and they kept it closed. During the 1996
Performance Test the exit was another door they kept closed (Appendix, Exhibit 2.1).

4.6 Start-up Features

       The Regenerative Thermal Oxidizer, or RTO, is an add-on control device manufactured by
Durr.  It has three towers and a "Programmable Logic  Controller" (controller) that operates the
system. The operator simply pushes the appropriate  button to start it or shut it down.

       The controller checks the RTO's operating requirements and verifies that all damper valves
operate from fully-open to fully-closed.  The controller also bars process air from entering until the
combustion chamber achieves the set point temperature and lights the burner.

       The RTO has a diagnostic program that can detect and display failures of the control system and
failures of field equipment. The controller detects these failures and causes an alarm to sound and a
rotating beacon to flash. The alarm continues until an operator pushes a "silence alarm" button. The
beacon stays lit until the fault is corrected.

       MMC is likely to move the CAPE+RTO System from one dock to another. After moving it to a
new dry dock but before delivering process air to the RTO an operator must:

       (a) Check for damage;
       (b) Ensure that all  piping, electrical, and duct connections have been properly assembled: and
       (c) Maintain the minimum required vacuum.

4.7 Operational Aspects

       Spray blasting (hull cleaning) and applying a coating can damage the flexible elements on the
CAPE's curtains or towers, but following proper operational procedures should minimize such
occurrences.

       Moving the paint-mix containers inside the CAPE before applying a coating reduces the need to
open the door during a coating cycle. With proper planning workers would only need to open it briefly
while entering and exiting the CAPE.

       Opening a door may drop the vacuum inside the CAPE below the minimum EPA Method 204
requires [6].   But MMC can make sure that doors to an enclosure stay closed while applying coatings.

                                            4-4
-------
For example, some spray booths have on the doors of the coating enclosure instruments that record the
number of times the doors opened and how long a door is kept open. In other facilities alarms indicate
when a door is open or when the vacuum inside the enclosure does not meet the minimum. MMC can
use these options if necessary, but we will not require them.

       After the RTO is fully operational and VOC or VOHAP-laden air is feeding into it, someone
must verify that it is operating properly including burner's fuel-to-air ratio. Here is the procedure:

       (a) Calibrate a portable, hand held, flame ionization detector using a gas standard with 0, 250,
and 500 ppmv of propane.
       (b) Use the detector to measure the VOC or VOHAP concentration at the outlet.
       (c) Check to see that the outlet concentration is 30 ppmv (as propane), or more above
background (5-20 ppmv).
       (d) If the concentration is too high and there is no readily assignable cause, call  a
manufacturer's representative to correct the problem and restore the RTO's efficiency.

       The RTO has a continuous recording device that monitors temperature in the combustion
chamber to keep it near  788°C (1450°F)--the temperature at which it passed its last performance test.  If
the combustion temperature drops below 760 °C (MOOT), the controller sounds a klaxon, triggers a
flashing strobe light, and diverts air from the CAPE directly to the atmosphere.  In this case, someone
will tell painters within the CAPE to stop operations.  MMC will keep  a copy of this written procedure
on file.

       The RTO can foul its ceramic heat-transfer beds if material coming to it does not burn up,
builds up on the beds, and causes the pressure drop across the bed  to increase. If the operator lets the
pressure drop to rise, it may choke the air flow and cause equipment problems upstream. Therefore,
the operating procedures must address this issue, typically based on advice from the manufacturer who
custom designed the RTO.

REFERENCES

1.  Application for Approval of Alternative Methodology for Compliance with The KESHAPfor
Shipbuilding and Ship Repair: submitted by Metro Machine Corporation. Norfolk. VA.  June 12, 1996
(Revised  October 31, 1996); prepared with Pacific Environmental  Services, Inc., Herndon, VA.

2.  Implementation Plan for Compliance \\'ith the NESHAP for Shipbuilding and Ship Repair Metro
machine Corporation; prepared  by Eric Lasalle, November 1, 1996; Metro Machine Corporation,
Norfolk, VA.

3.  Air Emission Evaluation Total Gaseous Organic Compounds and Filterable Paniculate Emissions
Compliant All Position Enclosure  (CAPE) System  USS SCOTT DDG-995 Metro Machine
Corporation; prepared by Pacific Environmental Services, Inc., Herndon, VA, September  1996  for
Metro Machine Corporation, Norfolk. VA.

                                           4-5
-------
4.  Clyde Smith, Mobile Zone Designsfor Ultra-Efficient Surface Coating Operations, Technical
Paper, Revised 11/89. Paper presented at the Air Pollution Control Association 1988 Congress and the
1988 Aerospace Symposium sponsored by EPA Region IX.

5.  Preparation, Adoption, and Submittal of State Implementation Plans; Appendix M, Test Methods,
204 A-204 F; 62 FR 32500 (Monday, June 16, 1997).
                                            4-6
-------
Appendix
  Exhibits
-------
Exhibit 2.1 CAPE side view showing one door no other openings
           (MMC's 1996 Emission Test)
•ACIFIC ENVIRONMENTAL SERVICES, INC.
                                Project No.
                                    5ie>~7. 001
                                              Page
                                               of
                                                            2.
                                Client
                                Location
Prepared By
^>
              Date
Checked By
Date
Sheet Title
                         - UL65
                                                               - Port
       -o
                               3D
                                       E-l
-------
Exhibit 2.2 Dimensions of the CAPE sections
            (MMC's 1996 Emission Test)
       PACIRC ENVIRONMENTAL SERVICES, INC.
                                    Client
                                     A
                                     Location
       Prepared By     Date  Checked By  Date  Sheet Title
       0^W_U) J L	i 15 *_
   '- x 6.  . =
5 i ' * 2   =
                                B.6.
                                           E-2
-------
Exhibit 2.3 Static pressure sampling locations
           (Fig. 3.3 in MMC's 1996 Emission Test)
J

- '
5= I
,

fN
55

§
•
i
F
,
a

-. o |» a

o

o
\±.







                                                                      —
                                                                     
-------
Exhibit 2.4  Proposed monitoring protocol
             (MMC's 1996 Application for Approval)
     PROPOSED MONITORING PROTOCOL

     MMC proposes to do the following to demonstrate compliance with the NESHAP using the
     CAPE® system.

           1.   Conduct a performance test to determine capture and destruction efficiencies of the
                proposed CAPE system.  A performance test protocol is given in Attachment I.
                The capture verification will demonstrate 100 percent capture and the VOC
                destruction efficiency test will show a minimum of 95 percent destruction.

           2.   Using the material balance approach described earlier and an overall control
                efficiency of 95 percent,  calculate the emission rate in grams of VOC per liter of
                solids for each application and compare  with the standard for each category of
                coating used.

           3.   Maintain and monitor the RTO chamber temperature at or above the temperature
                required for 95 percent destruction efficiency.  The desired minimum temperature
                will be established in consultation with the oxidizer manufacturer.

           4.   Ensure that all doors and windows to the enclosure are always kept  closed to
                capture VOC emissions.

           5.   Perform periodic checks  for any tears in the CAPE enclosure and repair as soon as
                possible.

           6.   Ensure that the RTO is on and operating properly for a period of time commen-
                surate with the type of coating being used.

     RECORDKEEPING AND REPORTING

        To determine and demonstrate compliance with the NESHAP, MMC will keep records of
     usage of all coatings, thinners, reactors and solvents for each application.  For each application
     of non-compliant coating,  the beginning and ending time of the coating application and the
     RTO operation will be recorded.  These records will be used to compile a monthly compliance
     record.  The compliance record will show the type of coatings used for each application during
     the month, the duration of each application and the number of hours of operation of the RTO
     for each application. The  records of monthly compliance determination will be retained for a
     period of five years and made available to the regulatory agencies upon request.
                                              E-4
-------
Exhibit 2.5 Example of protective coating quality assurance record for
             critical coated areas, Sheet 3 (MMC's 1996 Emission Test)
                                         Mal
                                                     Com
                                Protective CoAttnat Oualrrv Aoimn
D  Pre-Coat ( Stripe )  D  Disturbed Areas / Touch-Up
                                                     Application Method / Equipment
                                                                 Tim. Between Coats Of Paint:
      Dry Bulb
      Wet Bulb
                                                       67'F
      Relative Humidity
      Surfece Temperetura
                           «F
                                                                        -?#'/=
      Dew Point
      VI
                                MviutaCtuTvr
                                               Settalf
                                                                     Cal.
                                                                 //-//-9S
                                                                                 Due DM
      Surfeoe Temp.
                                            /47/S-l
      OPT. Gage
                        f/<.-,,~
                                                                                     - 7-17
      Profile Ge0«
                                            9'3CSO
      VII I Coating Inspection
                         Coat*
Sit
Unsat
1
(1) Cleanliness
(2) Visual Appeerance / Workmanship
(3) HoHdavs
(4) Ccrrection Of Defects
(5) WFT
(6) DFT
(7) Total Average DFT
Rfloulmd ^ - J»-t ^"
Actual •^
Reouind f-£.-,,}s
Actual (,.^^;\-,
Raouimd V~&~,,/f
Actual CT-*«I'K
(0) Drying And Curing Times Prior To Service
\/
V/
%/'
v/
vy
V/ ^
y #
v/








                                                                     Notet:
                                                                    1. Std Item 009-32 environmental
                                                                    reading* to be taken at a minimum of
                                                                    every 2 Hn.

                                                                    2. Painting on lurrace temperature
                                                                    below 3S*F or above 93*F requlrei the
                                                                    •pedflc approval of the Supervisor.

                                                                    3. Ambient and luroMe temperature
                                                                    mutt be at least 5*F above dew point.

                                                                    4. Relative humidity must not exceed
                                                                    85%.
     Remarkt / Comment*:
     Inspector Signature
                                                E-5
                                                       Date.
-------
Exhibit 2.6 Example of enclosure log
              (MMC's 1996 Emission Test)
                                             Enclosure Log
                                  77
Manometer Readings (Base)
Manometer Readings (Top)
Dry Bulb Temp (Outside)
Wet Bulb Temp (Outside)       7D
Dry Bulb Temp (Inside-Near middle of enclosure)
Wet Bulb Temp (Inside-Near middle of enclosure)
Estimate of % of enclosure at Negative Pressure
Are inflatable seals properly sealing?
Non inflated seals	
                                                     •07.   .0)
     Are there leaks/perforations in the.containment top?
     Is there a buildup of water on the ship's deck?
     Is the blast/paint foreman aware of any problems with the CAPE Towers?
     Are all of the towers functioning properly?
     Is each manlift functioning properly?
     Is water leaking from the ship?
     Is there evidence of water leaking into the enclosure?
     Is there evidence of water on the dock floor?       •     •-
                                                                                 Date
                                                                                Time  /
                                                                                            df foe  rz>i>:
      C:\OPERATE\LOGS\ENCLOSURE.DOC
      RCG
      08/16/96 10:40 AM
                                                E-6
-------
23/1998  16:12    919-785-9631 _                BERRY ENVIRONMENTAL                  PAGE  01

             Exhibit 2.7 Fax letter dated October 10,1998 from Berry Environmental to
                          U.S. EPA/OAQPS, in response to questions from EPA (2 pages)
   BERRY ENVIRONMENTAL
                                                                        P.O. Box 20634
                                                                        Raleigh, N.C. 27619
                                                                        Phone/Fax 919 785 963]

   Dr. Mohamed Serageldin
   Emission Standards Division, MD 13                                     Fax: 541 5689
   Office of Air Quality Planning and Standards
   U.S. Environmental Protection Agency
   Research Triangle Park, NC 27711

   Dear Mohamed

       This is in response to the questions that you asked in your Fax of October 19* regarding the statistics
   of the first "paint cycle" on October 20*, 1996 during the Performance Test of the CAPE System. Based
   on a discussion Jim McMichael of Metro Machine, some of the values should be changed.

   1. Amount of Paint Applied: 135 gallons
       Reference: The third page (with "Sheet 2" in upper right corner) of the several Operator log-sheets
   titled "Protective Coatings Quality Assurance Record for Critical Coated Areas" located in Appendix D
   of the Test Report. Notice on the log sheet that the mixing ratio is 4:1  with a maximum thinning of 5%.
   1 am told that it is unusual for the coating to require thinning so we cannot be certain that any was added.
       The reference that I believe you used, a computer-generated log-sheet located much later in
   Appendix D,  is confusing to me. Since it: 1) appears to have been created by PES as part of the
   Performance Test data and 2) does not reflect the appropriate 4:1 mixing ratio, we think  it is an inferior
   reference.

   1A. Time and length of paint cycle: 2 hours and 45 minutes, from 1815 hours to 2100  hours
       Reference: The fifth page (with "Sheet 3" in upper right corner) of the several pages referenced in
   item 1. above. That page is improperly dated at the bottom of the page (8-22-96) but properly dated at
   the top where the  start and stop dates and times are entered.)
       Note that the starting date of 1815 hours is at least 15 minutes earlier than the first inlet
   concentration data that is recorded on the first page of Appendix B-2 (EPA Method 25 A Data).

   2. Booth volume: Approximately 225,000 cubic feet or 6,371 cubic meters
       Reference: Letter1 dated July 26, 1996 from Charles Garland, Vice President of Metro Machine to
   James Cashel. As you know, the ship's hull, which constitutes one wall of the enclosure, is of complex
   curvature, so the calculation is an estimate of the actual volume.  Using the conversion factor of 0.02812
   cubic meters  per cubic foot, that calculates to a volume of 6,371  cubic meters.

   3. Actual flow rate to the RTO: About 12,595 cubic feet per minute or 21,420 cubic meters per hour.
       Reference: Calculated as the average of the first two entries of Table 2.3 on page 2-5 of the
   Performance  Test Report. Page 2-4 of the Test Report notes the  average flow rate during the test was
            This lener provides another interesting statistic, noting that a typical hull will require 1,500 gallons of paint
   compared to less than 200 for the superstructure area. Clearly the CAPE System will dramatically reduce the totaJ emissions
   from ship repainting operations.

           2 Actual! v, 0.0283)6847 according to the "Environmental Pocket Reference" by Mostardi-Plan Associates.

                                            E-7
-------
IS/23/1988   16:12
919-785-9631
                                                         BERRY ENVIRONMENTAL
                                                                                                  PAGE   02
        Exhibit 2.7 (Page 2/2)
      12,755 cubic feet per minute or 21,660 cubic meters per hour. The maximum flow rate during the
      performance test was 13,2783 cubic feet per minute (22,560 cubic meters per hour).
          Mohamed, based on this data, (which 1 had previously overlooked), please change Section C(3) of
      your final report.  Under the part titled "Flow through the RTO", please change the "Indicator Range" so
      that the upper limit is 14,000 cfm (ten percent greater than the average flow during the performance test).
       The lower limit should remain at 9,000 cfm.

      4. Spraying Rate: 49.1 gallons or 185.81* liters per hour.
          Reference: See items 1 and 1A above.

      5. Surface area coated: Approximately 13,000 square feet or 1,208; square meters.
          Reference: See upper right corner of references for items 1. and 1A. above. This first paint cycle is
      applied to a hull that has just been grit blasted. Because the entire hull uses a common prime coat, during
      the first paint cycle, paint is applied to the entire portion of the hull within the CAPE. This ^s indicated
      by the note in the Remarks section of the reference for item 1A which states that the 135 gallons were
      used to paint the "sides, underwater hull and seachest"

          Mohamed, I hope this information is helpful to you.

                                               Best Wishes
       cc: Rick Colver
              3 As you know, the RTO has a design flow of 11,000 cftn This test demonstrated the excellent "turndown" (or
       "turnup"?) ratio flexibility of this RTO design. Its efficiency averaged 99 percent even with when the inlet flow averaged 16
       percent above design (which, of course, decreased residence time in the combustion chamber a like amount).

              4 A conversion factor of 3.785 liters per gallon was used.
                A conversion factor of 0.0929 was used.
                                                   E-8
-------
Exhibit 2.8 Calculation of required control efficiency
           (Table 5 of MMC's 1996 Application for Approval)
              I
              W
              w
              u






10
u:
cc




to
Z
O
2
u
o
o
H
u
1
y
E
u.
UJ
              8
              CX
              g
              u.
              1/5
              §

              •J
              U






















1
o

U
N
y
k!
II
z 2
oy
ILITARY 1
S




z

D
U.
5




S.
bfi
C
8
-
s

s

55
O

O

cs
o
oo
S
o
vo



0
o

o
s
IS
C ,
U I
s 1
0
in
(N
s

*

o
m
S

P

=0
o
(N
0
o
IO



o
VO
O

o
oc
O <4- 1
u O —
•"
-• 1

to ^5
« 8

" s
- s
*—•'
5 S
^
- 1
Tf 00
~
oo NO
r-t m
10 S
« 9



3 1
O1- OO
TT vo
w
>/-> sc
a
g -| s.
uT •*. -a *:
< ° .£ -S
5 £ "E.
O s §•
-
0

S
o

5?
o
o
o
o
o
o
o
o
o
s
o
o
IO
o



o
C)
o
o
0
o
VS. = Volume ratio
of solids^} applied
coatine
10
p. p
•^j- to

S I

S 1
. t^*
!0
^*1 r**

r-
o
o

o
0

VO
10

-

X
o
0
o
0
ON



(S
ON
m
ON
S
(S
t~
OCE= Overall
Control Efficiency
Required (%) \
oo
o
£

p
V)
ON

O
vn
ON
p
in
ON
p
in
O
in
ON
p
ON
0
in
p
in
ON



O
in
0
in
ON
p .
in /
ON /
CAPE* Overall
Control Efficiency
CK}
ON
                                        E-9
-------
Exhibit 2.9  Number of RTO hours for compliance with the NESHAP
           (Table 2 of MMC's 1996 Application for Approval)
                                           TABLE 2

                    NUMBER OF RTO HOURS'REQUIRED FOR COMPLIANCE
TYPE OF COATING
(STANDARD)
ANTIFOULANT
[6.38 Ib VOHAP/gal (765
g VOHAP/L)solids]
GENERAL USE -
Inorganic Zinc, Military
Exterior [4.76 Ib VOHAP
/gal (571 g VOHAP/L
solids)]
SOLIDS CONTENT
OF COATING AS
APPLIED
(% volume)

20
30
40
50
60
80
20
30
40
50
60
80
HOURS OF RTO OPERATION FOR
COMPLIANCE6
2 Hours
4
3
2
1
0
0
4
3
2
2
0
0
4 Hours
5
4
3
2
0
0
6
5
4
3
6 Hours
7
6
4
2
0
0
7
6
5
4
0 0
8 Hours
8
7
5
2
0
0
9
8
6
5
0
0 00
          4CAPE SYSTEM EVALUATION AND EMISSION TEST REPORT at MMC Compliance
          Engineering, Inc., Philadelphia, PA. December 14-21, 1995, Durr Project No. 2996-1021

          5AIR EMISSION EVALUATION TOTAL GASEOUS ORGANIC COMPOUNDS AND
          FILTERABLE PARTICULATE EMISSIONS at Metro Machine Corporation, Norfolk, VA, August
          19-23, 1996, PES Report 5187-001.

          6Rounded up to the next integer number
                                      E-10
-------
Exhibit 2.10 Minimum number of hours of RTO operation required to achieve
           background level (Table 4 of MMC's 1996 Application for Approval)
                                        TABLE 4

                 MINIMUM NUMBER OF HOURS OF RTO OPERATION REQUIRED TO
                    ACHIEVE CONCENTRATIONS OF LESS THAN 908 g (2.00 Ibs)'
                 FOR VOHAP APPLICATION RATE OF 160 LBS/HOUR (72,640 g/HOUR)
COATING TYPE
ANTIFOULANT
HIGH VOHAP
CONTENT [6.381b/gal
(765 g/L) solids]
GENERAL USE -
Inorganic Zinc, Military
Exterior LOW VOHAP
CONTENT [4.76 Ib/gal
(571 g/L) solids]
SOLIDS
CONTENT OF
COATING AS
APPLIED
(% by volume)
20
30
40
60
50
80
20
30
40
50
60
80
TOTAL HOURS OF RTO OPERATION
TO ACHIEVE BACKGROUND
CONCENTRATION
2 Hours
Coating
8
8
8
8
8
8
8
8
8
8
8
8
4 Hours
Coating
10
10
10
10
10
10
10
10
10
10
10
10
6 Hours
Coating
12
* 12
12
12
12
12
T 12
12
12
, 12
. 12
12
8 Hours
Coating
14
14
14
14
14
14
14
14
14
14
14
14
                                   E-ll
-------
11/86/1997  12:24    919-765-9631                BERRY ENVIRONMENTAL                PAGE  81

      Exhibit 2.11  Fax letter dated November 6,1997 from Berry Environmental to
                     U.S. EPA/OAQPS, in response to questions from EPA regarding
                     dry to touch time (1 page)
       BERRY ENVIRONMENTAL                                                     	
                                                                    P.O. Box 20634
                                                                    Raleigh, N.C. 27619
                                                                    Phone/Fax 919 785 9631

                                                                    Date: November 6, 1997
           FAX COVER SHEET
           To: Mohamed Serageldin

           Dear Mohamed

                 Last week you asked for estimates of the amount of VOC remaining in a marine coating whe.
           becomes dry to touch. In response you received estimates from two analytical chemist;'

           Hiro Fujimoto, formerly of BASF                       About 3 percent
           William Golton, formerly of DuPont                     Less than 10 percent

                 The additional information that we discussed earlier today is presented below. It was obtaine
           by visiting booths of several exhibitors at the International Coating Exhibition earlier this week in th
           Atlanta convention center. At each booth I asked for the most knowledgeable epoxy forrnulation
           chemist available. Each was then asked for an estimate of the percent of VOC remaining when a spr
           applied, thick film (3 to 4 mils), air-dried, epoxy-amine coating has "dried to touch". None was
           anxious to answer, noting that there were a lot of variables (solvent type, mix, oligomer length and re
           type were named) that could affect the rate  of evolution. When it was explained that no single coatir
           was involved, rather the quest was for a reasonable guesstimate for a variety of marine hull coatings,
           following values were given.

           Richard Martorano, Rohm and Haas,                    5 to 6 weight percent
           Larry Wang, Reichhold                                Less than 10 percent
           Andy Wang, Ciba                                    15 to 20 percent

                 Neil Wassburg (409 238 4420), was identified as best qualified to answer the question at Dov
           booth, but he was never there when I was.

                 If all of the  estimates are averaged with tire "less than 10" scored as a full 10 percent, the
           average is 9.1 weight percent. If we score this like in ice figure skating, throw out the high and lew, i
           average would be something even less than 9 weight percent.

                 Mohamed, I hope this helps in your decision-making. I look forward to seeing you next week
-------
Exhibit 3.1  MSDS for Intergard FP showing drying time (hours)
 Intergard*  FP
 Polyamine   Epoxy
                                                          Intemationc
                                                          Marine Coatings
 INTENDED Us»
 A universal anncorrosive for use on underwater hulls, above water areas, internal areas including
 ballast tanks, marine vessels, barges and offshore structures.
 P«ODUCT DuuirTioN     A rwo component, self priming, surface tolerant epoxy with semi-gloss finish. Exhibits excellent
                       chemical and abrasion resistance. A high build formulation capable of low temperature
                       cure. Low VOC,
 PRODUCT INTOKMATION
Color

Finish/Sheen
Convener
Volume Solids
Mix Ratio
Flash Point

Film Thickness
(SSPC-PA2)
White - FPD052: Light Gray - FPJ034: Red - FPL274. Special colors can o.
matched to meet customer specifications.
Semi-Gloss (ASTM D-523)
FPA327 for normal applicauons/FCA321 for low temperature
30.0°t = 2°c (ASTM D-2697) at 77'F (25'C) and 7 davs cure
4:1 bv volume
Part \. 1 17°F <47'O: Pan B. 124'F ol 'Ci. Mixed.  1 17'F >47'C;
. Seta/lash . , .ASTM D-3273<
4 H mils dn. -.Decided equivalent to j 0 mils ^vet
4 iRi.O mils arv practical range equivalent to 5 i)-7 3 mils wet
                      Theoretical Coverage      320 sq. ft.  ?al (4.0 mils DFT) Allow appropriate loss factors
 DETAILS
Method                Comenuonal or airless spra%
Induction/Sweat-in Tone  15 minutes it temperatures belou 70'F i21 '(Ii
Thinner                GTA415 See Regulate rv Data
Cleaner                CT A 415
Pot Life                •< hrs 4 ?<) F • 10 C"i  4  hrs 4 "5 F "J4 f   -' hrs 4 9
-------
Exhibit 3.2   CAPE and RTO layout
             (MMC 1998 Communication)
  60,000 VOC
   Laden Air
                                      Exhaust
                                      1.1 ,000 Clean Air
                                                               	I,
                                                  24,5000 / Fan
                                                                       3
   49,000
VOC Laden Air.
                                    CAPE Enclosure
                                             11,000 CFM Infiltration
                         C.A.P.F.   PAINT  MODE
Reference
C.Garland and M. Lukey; An Innovative Permanent Total Enclosure for Blast Cleaning and Painting
of Ships in Drydock, MMC Compliance Engineering, Inc.
                                        E-14
-------
Exhibit 3.3  Sample calculation for 8 hour paint application
              (MMC's 1996 Application for Approval)
                               SAMPLE CALCULATION FOR
                               8 HOUR PAINT APPLICATION
    Maximum duration of coating
    VOC application rate
    VOC density
    Solids content by volume
    Solids volume as applied
    Total solids volume
    Total VOC applied
   8 hr
 160 Ib/hr
 7.0 Ib/gal
  20 percent
5.71 gal/nr
45.7 gal
1280 Ibs
 72.640 g/hr
   840 g/L

   21.6 Uhr
   173 L
581,120 g
     Inside CAPE enclosure
     VOC emitted
     Total VOC emissions

     Volume exhausted thru RTO
     Volume circulated thru CAPE
     Overall Control efficiency
96.9 Ibs
62.0 Ibs
159.0 Ibs
11,000 cfm
60,000 cfm
95%
44,011 g
28,162 g
72.172 g
312 cmm
1,700 cmm
     CALCULATION FOR COMPLIANCE WITH 4.76 lb/aai-{571 g/LJ solids
Hours

1
2
3
4
5
6
7
8
9
VOC added
(Ibs)
160
160
160
160
160
160
160
160
0.00
TOTALS* | 1,280
(a)
72,640
72.840
72,640
72.640
72.640
72.640
72,640
72.640
0.00
S81.120
voc witnin CAHC
(Ibs)
121
175
199
210
214
217
217
218
96.9

W
54,998
79,464
90,351
95,190
97,347
98,305
98,731
98,922
44,011

VOC to RTO
(Ibs)
61 .1
116
140
151
158
158
159
160
139
1,241
(9)
27,726
52,659
63,751
68,686
70,878
71,859
72,290
72,486
62,979
563,314
VOC Emitted
(Ibs)
3.05
5.80
7.02
7.56
7.81
7.91
7.96
7.98
6.94
52.0
(g)
1,385
2,633
3,187
3,432
3,546
3,591
3,614
3,623
3,151
28, 1 62
Status of
CAPE
ON
ON
ON
ON
ON
ON
ON
ON
ON

                                                                                               it/
     Number of RTO hours for compliance = 9
                                                 VOC emitted 3.47 Ib/gal (416 g/L) solids
     CALCULATIONS TO ACHIEVE A CONCENTRATION OF LESS THAN 908 g (2.00 Ibs)
Hours
1
2
3
4
5
6
7
8
9
10
11
12
13
14
VOC added
Ibs)
160.00
160.0Q
160.00
160.00
160.00
160.00
160.00
160.00
0.00
0.00
0.00
000
0.00
0.00
(9)
72.640
72.640 +
72,640
72,640
72.640
72,640
72,640
72,640
0.00
0.00
0.00
0.00
0.00
0.00
VOC within CAPE
(Ibs)
121
175
199
210
214
217
217
218
96.9
43.1
19.2
8.54
3.80
1.69
(9)
54,998
79,464
90,351
95,190
97.347
98,305
98,731
98,922
44,011
19,581
8,712
3,877
1,725
767
VOC to RTO
(Ibs)
61 .1
116
140
151
156
158
159
160
139
44.0
196
8.70
3.87
1 72
(a)
27,726
52,659
63,751
68,686
70,878
71,859
72,290
72,486
62,979
19,953
8,876
3,950
1,757
781
VOC Emitted
(Ibs)
3.05
5.80
7.02
7.56
7.81
7.91
7.96
7.98
6.94
2.20
0.978
0.435
0.193
0.086
(g)
1,385
2,633
3,187
' 3,432
3,546
3,591
3,614
3.623
3,151
999
444
197
87.6
39.0
Status of
CAPE
ON
ON
ON
ON
ON
ON
ON
ON
ON
ON
ON
ON
ON

      Number of RTO hours to achieve a concentration of less than 908 g (2.00 Ibs) = 14
                                              E-15
-------
Exhibit 3.4 Paint system for Scott (DDG - 995)
         (MMC's 1996 Emission Test)
          y8228R
          S |S 6 e> 6 6
            *>  to •
                                   §333;;
                                 e
                                 I
                                 j

                                 g
   3
   o
    O
    •
    o
   J

SJi'
Si5
                  i
                  s
      e '
     o t- i
                1
      * 0<
      35
SRS
IZMI
                           8 8
          M;
          si*
                        n !
                        i
                        ft
                        t
                        £

                            N
                           n •
                         1 1!
                         I?--
                         n

                                     -
    i
    n
    ^


    I
                             il
        I
        \
        I
    mi
    m
   I
   i
  I*-"
         I
         1
                         111]
                      8
                      «
                      3
111
I!!
                             E-16
-------
Exhibit 3.5  Example of protective coating quality assurance record for critical coated
    -'       areas, Sheet 2 (MMC's 1996 Emission Test)

     Location.
                                          Metro Machine Com
                                 Protective Coatings Quality Aasuranea
                                        For Critical Coated Areas
                                      /3,^00 i<»
                       D  Pre-Coat (Stripe)
                                      Coat*
                          O Disturbed Areas / Touch-Up

                                   Sat         Unsat
r ~r /
\ (1) Paint Manufacturer Name -* /^ />*'" **> *e 7"/"><*-
f
$y Untat
t/.
tX x
v/
      Remarks / Comments:.
                        G*//.
            I35
      Inspector Signature_
                                             E-17
-------
Exhibit 4.1.  Plan showing CAPE modular units and barge carrying the incinerator
                 (Barge Transporting RTO  and

                 Air Management  System)
                                       TOLERANCE TABLE
                                          10UUNCCS . !
-------
Exhibit 4.2 Summary of stack gas conditions
            (Table 2.3 in 1996 Emission Test)
                to
                  W
            rn
            fS
            u   u:
            I—I
            pa
            g
^<=^
2 J w
S < ^
2s^
O
5R
                    u
                  K
                  O
                           o"*
           O ;
           U
                            H
           u^a
           II!
                                   ooooooooo
                                    ppppppppp"
                                    ddociodo'oci
                                    C.C.!
                                                 -
                                      i fO f*i f*"> rn f> P"I f)
                                    PT— $-
-------
                                    TECHNICAL REPORT DATA
                               (Please read Instructions on reverse before completing)
 \ REPORT NO
     EPA-453/R-99-005
                                                                  3 RECIPIENT'S ACCESSION NO
 4 TITLE AND SUBTITLE
    Evaluation of Application for Approval of an Alternative
    Methodology for Compliance with the NESHAP for
    Shipbuilding and Ship Repair and Recommended Requirements
    for Compliance (Application Submitted by Metro Machine
    Corporation, Norfolk, Virginia)	
                 5 REPORT DATE
                   July 1999
                 6 PERFORMING ORGANIZATION CODE
   AUTHOR(S)
    Mohamed A. Serageldin, Ph.D.
                                                                  8 PERFORMING ORGANIZATION REPORT NO
 9 PERFORMING ORGANIZATION NAME AND ADDRESS
   U.S. Environmental Protection Agency
   Office of Air Quality  Planning and Standards
   Research Triangle Park, NC  27711
                                                                  10 PROGRAM ELEMENT NO
                 11 CONTRACT/GRANT NO
 12  SPONSORING AGENCY NAME AND ADDRESS
    Director
    Office of Air Quality Planning and Standards
    Office of Air and Radiation
    U.S. Environmental Protection Agency
    Research Triangle Park. NC  27711
                 13 TYPE OF REPORT AND PERIOD COVERED
                   Final
                 14 SPONSORING AGENCY CODE
                   EPA/200/04
    SUPPLEMENTARY NOTES
    Project Manager: Mohamed A. Serageldin. Ph.D.. (919) 541-2379
 16  ABSTRACT
 The U.S. Environmental Protection Agency is providing background information that supports the use of
 Metro Machine Corporation's (MMC) compliant all position enclosure (CAPE) plus air management
 system and regenerative thermal oxidizer (RTO) (CAPE + RTO System) as an alternative means  of limiting
 the emissions of volatile organic hazardous air pollutants per volume of applied solids (nonvolitiles).  This
 document also explains how we arrived at the operating, recordkeeping. and reporting conditions that
 MMC must meet for approval. The add-on control system they used consists of a pollution capture unit
 operation (CAPE) plus air management system and a destruction unit operation (RTO). When operated
 according to the specified procedures, it will control emissions to a level no greater than that from using
 coatings which comply with the limits in Table 2 of 40 CFR Part 63, Subpart II.	^_^
                                      KEY WORDS AND DOCUMENT ANALYSIS
                   DESCRIPTORS
                                                 b IDENTIFIERS'OPEN ENDED TERMS
                                                                                    c COSATI Field/Group
 Air pollution, Equivalency, Shipbuilding and
 Ship Repair, NESHAP, Marine Coatings,
 Surface Coating, Ship Painting (Coating)
Air Pollution control
  18 DISTRIBUTION STATEMENT


    Release Unlimited
19 SECURITY CLASS (Repon)
   Unclassified
21 NO OF PAGES
       81
                                                 20 SECURITY CLASS (Page)
                                                    Unclassified
                                                                                    22 PRICE
EPA Form 2220-1 (Rev. 4-77}   PREVIOUS EDITION IS OBSOLETE
-------
y S  Environmental Protection Agency

Region 5, Library (PL-12J)    9 . fto
77 West Jackson Boulevard, 12U» ru
                  -
60604-3590
-------