DEQ-CFW_00072366

ULIN i twi, UPPIL-L F KIV111'FRAC:KING SLIP
Facility Name:_- _�JiJY GYtT
h u.vnt-ct.f—�e�zt-te
y I a
��1e Facility/Application ID: � C%f)C�C t
. tf A
County/Regional Office: to e V1 -y:ye
, Engineer: v( (s„ - G %V�
Send Regional Office Copy of Application: Yes /No ZeC cr',_e v-- Le%tc�
PART I - ACCEPTANCE CHECKLIST A4 F
Acknowledgement Letter: Already Sent
/Please Send
Initial Event(s): C TV-Ack./Complete
_: State Ack. Letter due
TV-Ack./Incomplete add info
C State App. not accepted — add info request
Fee Information:
Acceptance Check List:
Amount Due: PSD or NSR/NAA $13,488
Yes No
NA
PSD and NSR/NAA $26,235
Appropriate Number of Apps Submitted (minimum of 2)
TV Greenfield $8,910
Application Fee Submitted
TV $867
Zoning Addressed —
Ownership Change $62
Source Recycling/Reduction Form Submitted C
Renewal/Name Change — NA
Initial 7, rO
Authorized Signature
-=
Amount Received:
PE Seal
Additional Amount Due:
PART II - IBEAM UPDATES PART III - COMPLETENESS CHECKLIST
Application Type: Permit Application Schedule:
u Additional Permit ❑Appeal
0 Director Administrative Amendment ❑ Required Application Forms Submitted and Completed
C Administrative Amendment 0 Expedited State
❑ State 0 Supporting Materials & Calculations Received
0 Appeal 0 PSD
0 PE Seal (If 15A NCAC 2Q .0112)
E. Greenfield Facility
❑ Modeiing Protocol Acceptance
0 Last GACT/Toxics
0 Confirmation of Pollutants Modeled
0 Last MACT/Toxics n TV — State Only
Modification ❑ TV — Expedited
❑ TV — 502(b)(10) ❑ E5 Form (Significant Modifications)
2'1-V — Minor
0 Name Change 0 TV — Greenfield
0 TV — Renewal
0 New Permit --'TV — Reopen for Cause
0 TV — Significant (2Q .0501(c)(2))
D Ownership Change 0 TV — Administrative
0 TV — Significant
0 Renewal 0 TV — Ownership Change
0 TV — I" Time
0 Renewal w/Modification
PART
IV - GENERAL COMMENTS
e t4t
.�.�t� C b e V-
PART V - SUPERVISOR REVIEW
TVEE Updated (by Engineer): TVEE Verified:
PART VI - CLOSEOUT INFORMATION
Regulations Applicable to This Application (indicate all new regulation :
❑ NESHAPS/MACT ❑ PSD/NSR Toxics/Combust'on Sources After 7/10/10
NESHAPS/GACT D/N`SR Avoidance D 1 SIP Regulations (list all new):
❑ NSPS ❑ Existing Source RACT/LAER
❑ 2D .l 100 ❑ New Source RACT/LAER
❑ 2Q .0711 ❑ RACT/LAER Added Fee
O 2Q .0705 Last MACT/Toxics ❑ RACT Avoidance
HAP Major Status (after) 21 ajor 0 Minor
PSD or NSR Status (after) ❑ Major for
Miscellaneous 0 Multiple Pen -nits at Facility
Not Determined
❑ Multi -Site Permit 0 Recycled Oil Condition
Dates Issue Effective
IBEAM Closed Out By: Permit Number:
❑ Public Notice Published -_ Public Notice Affidavit Of not noticed via DAQ Website)
Document Manager Updated by Engineer: Date:
,5��n,A-1,io 2-
Tracking- Slip v36 - mjc
Expiration
Permit Class Information
Before Auer
Small the V
C Syn Minor
�e V
0 Proh Small
0 General
0 Transportation
Revision Number:
DEQ-CFW 00072366

DuPont Fluoroproducts
Fayetteville Works Plant
22828 NC Highway 87 West
Fayetteville, NC 2830E-7332
VIA ELECTRONIC SUBMITTAL
August 16, 2011
Mr. Mike Gordon
NCDENR — Division of Air Quality
Permitting Section
1641 Mail Service Center
Raleigh, NC 27699-1641
SUBJECT: Additional Information Regarding Accuracy of Micro Motion Densitometer
Title V Permit Modification Application ID No. 0900009.11A
Title V Air Permit No. 03735T36
Dear Mr. Gordon:
This note is the requested response regarding the validation of the accuracy of the densitometers
that are being used to determine the concentration of vinyl fluoride ("VF") in the Maintenance
Header Vents (Emission Source ID Nos. FEP-112 and FEP-C2).
The permit application's Form D5 states that the densitometers' performance has been validated
and has been determined to have an accuracy and precision of two percent of full scale or better.
The source of that statement is a combination of specifications from instrument vendors and an
equation that was derived by area personnel that equates average molecular weight of a mixture
of VF and nitrogen to the concentration of VF in that mixture expressed as a percentage.
A .pdf file of the technical specification sheet for the Emerson Micro Motion 7812 Gas Density
Meter ("densitometer") is being sent to you separately due to the size of the file. On Page 3 of
that document, Emerson states that the 7812 Gas Density Transducer's density accuracy is better
than +/- 0.10% of reading for nitrogen and better than +/- 0.15% of reading for natural gas and
ethylene. For our purpose, DuPont used the +/- 0.15% of reading as the assumed accuracy for
VF.
The 7812 densitometer has a built-in temperature probe that is used for the determination of the
gas density as well as for the percentage VF calculation (discussed below). On Page 3 of the
7812 technical specification sheet, Emerson states that the 7812 Gas Density Transducer's
temperature accuracy is better than 0.5°C, which is just under 0.5% of range.
E.I. du Pont de Nemours and Company
DEQ-CFW 00072367

Y1r. Mike Gordon
NCDENR — DAQ
August 16, 201 I
Page 2 of 3
A .pdf file of the reference manual for the Rosemount 3051 Pressure Transmitter is being sent to
you separately due to the size of the file. In Attachment A, Rosemount states that the 3051T
Range 3 Specification's pressure accuracy is better than +/- 0.065% of span.
The first step in determining the concentration of VF in each Maintenance Header Vent at any
given moment is to determine the average molecular weight of the gas, which again is
exclusively a mixture of VF and nitrogen. This is done using the Ideal Gas Law and the
measured parameters of density (lb. of gas per cubic foot of gas), temperature (deg R), and
pressure (psia), and the gas constant of 10.7136 cubic feet - psia per deg R per lb -mole:
A4'W ��nt��=RxTxD.- P
Instrument manufacturers express the accuracy of their products at the 95% confidence interval
or two standard deviations of the measured values versus the true value. To determine the
overall accuracy of the measured average molecular weight the following steps were followed:
• Convert the individual accuracies of the density, temperature, and pressure measurements
from two standard deviations to one standard deviation.
• Square those results to find the individual variances of each of the measurements.
• Sum the variances and find the square root of that sum to determine the overall standard
deviation of the measured average molecular weight.
• Double that overall standard deviation to determine the desired accuracy of the measured
average molecular weight.
Measured
Parameter
Specified Accuracy
( 2cr )
Standard Deviation
( a )
Variance
(a2 )
Density
0.15 %
0.0008
5.63 (10-7 )
Temperature
0.50 %
0.0025
6.25 (10"6 )
Pressure
0.065 %
0.0003
1.06 (10'? )
Sum of variances
6.25( 10-6 )
Overall Variance ( a2)
6.25 (10"6 )
Overall Standard Deviation ( a)
0.00263
Overall Accuracy (2a)
0.53 %
Therefore, the overall accuracy of the measured average molecular weight of the gas stream in
the Maintenance Header Vents is estimated to be +/- 0.53% of the fiill scale.
DEQ-CFW 00072368

Mr. Mike Gordon
NCDENR —DAQ
August 16, 2011
Page 3 of 3
Area personnel have calculated the theoretical molecular weight of a mixture of VF and nitrogen
at increments of 10% VF. This work is shown in the Excel spreadsheet that is being sent to you
separately. A best -fit equation was then found for the plot of %VF versus Average Molecular
Weight. This equation, displayed in the spreadsheet's imbedded chart, is ultimately used to
calculate the fraction of the Maintenance Header Vent that is VF. In column C of the
spreadsheet is the percent VF that is predicted by the equation. Comparing the theoretical % VF
in Column B to the predicted % VF in Column C (excluding the values at 0% VF), the average
relative error was calculated to be 1.18%.
Considering the overall accuracy of the average molecular weight measurement of 0.53% and the
1.18% relative error of the best -fit VF concentration equation that uses the average molecular
weight measurement, DuPont believes that determining the VOC emissions (EFBrsca) from the
Maintenance Header Vents (Emission Source ID Nos. FEP-B2 and FEP-C2) using the densitometer
data will be accurate to within two percent of full scale or better.
If you need additional information, please call me at [phone] or send me an email.
Environmental Manager
cc: Gregory W. Reeves, NCDENR — DAQ, FRO (via electronic submittal)
John C. Evans, NC DENR — DAQ, Raleigh Central Office (via electronic submittal)
DEQ-CFW 00072369