I-DEV0437: Deviation from monitoring of the parameters, C_SF6,y and M_d,in

>>Description of noncompliance with the monitoring plan

This request for deviation is being submitted to address three issues related to monitoring C_SF6,y and Md,in that were raised during initial and first verification regarding emission reductions from the Samsung Electronics SF6 abatement project during the monitoring period from 21 October 2010 to 28 February 2011. These issues are discussed as below:

1. Monitoring of the consumption of SF6 (C_SF6,y)
As per the monitoring plan in the registered PDD of the project activity, C_SF6,y is annual consumption of SF6 during the year y, and it is defined as the total SF6 purchase in a specific project year y taking into account the change in inventory in the same year.
In the course of first verification, it is found that the consumption of SF6 in the LCD manufacture process was measured and recorded with direct measuring from the load cell installed at the site. Direct measuring with load cell is considered more accurate method to monitoring of C_SF6,y, whereas C_SF6,y shall be measured as purchase taking into account the change in inventory according to the monitoring plan in the registered PDD.

However, purchasing record, stock change data, and heel value are need to calculate C_SF6,y with inventory method. All purchasing records were identified with G-ERP system, but stock record at the start date of this monitoring period has not been verifiable with documented evidence. Accordingly assumptions for stock change and choice of heel value are needed in order to apply inventory method.
Due to above reason, the validation team can not give assurance for the compliance with the monitoring plan.
For the purpose of ensuring conservativeness in emission reductions, PP decided to use lower value between the results of direct measurement and inventory method even though direct measuring can give accurate value from the general point of view and to request for deviation from the monitoring plan for this monitoring period.

In the monitoring period, PP will apply C_SF6,y with inventory method which shows lower value. Details on each method are as below.

-The monitoring of C_SF6,y with inventory method
Purchasing amount during the monitoring period is 29.420 ton which is verified through purchasing record from G-ERP system.

For the stock, the record of the stock at the start date of this monitoring period was not verifiable with documented evidence. But through the gas room operating log, it is identified that there is no gap between the number of input record and output rocord of SF6 cylinder during the monitoring period. Taking into account that used cylinder (empty) is needless to store at the site, even the initial stock record was not verifiable, it can be assumed that there is no change in stock amount during the monitoring period.

As for the heel value, i.e. un-used SF6 gas left in the gas cylinder, 10% as IPCC default value is applied conservatively. As the actual heel value was measured as 9.91% based on the direct weight measuring records with load cell (the measured data is corrected according to EB 52/Annex 60”. Refer under delayed calibration history), applying the bigger value, 10% is more conservative for heel value.
Consequently, C_SF6,y is calculated applying assumption of no change in stock amount in the monitoring period and 10% heel gas ratio according to IPCC guidline.

The result of C_SF6,y with inventory method is 26.478 ton (29.420 ton * (1-10%) = 26.478 ton)

-The monitoring of C_SF6,y with direct measurement
For SF6 consumption in the monitoring period, 27.192 ton is measured using load cell with 99.9% accuracy(from the specification, it is identified that the accurary of installed load cell is 99.9% of F.S:Full Scale).

Considering other SF6 emission reduction methodology, AM0035 (“SF6 Emission Reduction in Electrical Grid”, version 01: this methodology requires 99% accuracy to monitoring SF6), this accuracy level is thought to be valid for this project. And for the QA/QC, the load cell needs to keep calibration as per manufacturers’ recommendations or national regulation.
But during physical site inspection, it is identified that the calibration of load cell had been delayed. Since installation of equipment on 24 November 2005, no calibration has been undertaken. Considering the regulation on measuring in Korea, this measuring equipment has to be calibrated every two years, the delayed period is about 4 years. The delayed calibration implemented on 24 October 2011 and the result shows that the error is 0.15% of F.S. This error is beyond the maximum permissible error (0.1% of F.S). According to “Guidelines for assessing compliance with the calibration frequency requirements (EB 52/Annex 60), the identified error is applied to correct the measuring value over the monitoring period. (27.192 ton * (1-0.15%) = 27.151 ton )

In summary, the C_SF6,y from direct measuring method is 27.151 ton SF6 and the C_SF6,y from inventory method is 26.478 ton SF6 and the lower value, 26.478 ton will be applied for C_SF6,y


2. Measuring and calibration method of SO2F2 pertaining to M_d,in
M_d,in is defined as total dry molecular weight of inlet stack gas. The averaged relative concentrations of SF6, Ar, He, O2, CO2, N2 and other gases with concentration of greater than 100ppmv, such as CF4, CO, HC, HF, SO2F2 SOF2 and SiF4 should be quantified with measurement by the QMS as per the methodology and registered PDD.
In the project activity, SF6, Ar, He, O2, CO2, and N2 are scanned as over 100ppmv gases in inlet stack and additionally only SO2F2 among other gases is scanned as over 100ppmv gas in inlet stack gas. Thus, SO2F2 also need to be measure by the QMS.
However, there is no certified standard gas for SO2F2 available because this gas is problematic with the stability of gases. It is confirmed through checking the existing of standard gas in the BIPM(Bureau International des Poids et Mesures) website (http://kcdb.bipm.org/appendixC/default.asp). If there is no standard gas, the QMS system could not measure the gas because QMS need to use calibration with standard gas for quantitative analysis.
Instead, SO2F2 can be analysed with FTIR and library information provided by the FTIR manufacturer.
The FTIR system has equal accuracy with QMS and it is suitable for gas analysis. The absorption intensity for FTIR follows the ‘Lambert Beer law’ which details the coefficient, path length, and number of molecules of a selected component. Using this law, without gas standards, it can be quantify certain components by FTIR with their known absorption coefficients. And many FTIR manufacturers provide a library which contains most useful infrared absorption wave-numbers with corresponding absorption coefficients for many gas chemicals, such as SiF4, and SO2F2.

To demonstrate the validity of FTIR calibration using library information instead of standard gas, related test result (This is the result of Dr. Jin Seog KIM’s study in KRISS) with object gases (SiF4) is reviewed. SiF4 is selected as object gas because it has standard gas and library information and is listed as one of other gases in M_d,in in the applied methodology (AM0078 version 01.1).

A known concentration of SiF4 is tested by FTIR system (OTSUKA, same manufacturer of FTIR used in the project) and quantified the concentration using library information. As a result, it is identified that the difference of known concentration and measured concentration will be within measurement uncertainty.
The following is an example of this quantification method for SiF4.
First, operator gets a standard gas mixture of SiF4 5000 ppmv in nitrogen which was prepared using gravimetric method by a national metrology institute and applies this standard to the calibration system. A range of SiF4 concentrations is generated by mixing a SiF4 standard (e.g. 5000 ppmv SiF4 in N2 ballast) with nitrogen, using the calibration system by two mass flow controllers. The pressure in the sampling line is measured and maintained at a constant slightly below atmospheric pressure using a pressure gauge and a needle valve to adjust the flow. The operator will record a measured concentration by FTIR with library information and obtain the relative difference between admitted concentration generated by calibration system and obtained concentration. The operator will ensure that a minimum of 5 points from 200 ppmv to 2000 ppmv should be measured to obtain relative differences. The average of absolute relative differences should be less than 5 % to meet the uncertainty described in this methodology. An example of the SiF4 quantification results measured by FTIR library method is presented in Table 1 and Figure 1. In this example, the relative standard uncertainty for using FTIR library method is less than 5 % as shown in Table 1 (refer to attached Table 1 and Figure 1)

KFQ reviewed relevant documents to check described above and interviewed chemical analyst (Dr. Jin Seog KIM (e-mail : jkim@kriss.re.kr) is a principal researcher of center for gas analysis in KRISS and a professor of University of seicence and technology in division of seicence of measurment) and manufacturer of QMS and FTIR. Their opinions are that the alternative measuring (FTIR use instead of QMS for SO2F2) and calibration method (calibration using library information when the standard gas does not exist) are valid and appropriate for SO2F2. The accuracy of measurement would be also improved. Thus, we confirm this alternative measuring can be more performable.
PP will correct the measured value applying 5% uncertainty for M_d,in in a conservative manner(From the test result, the maximum difference is 3.74% comparing between two measured results, but the 5% of uncertainty is applied to correct the data conservatively). The averaged mole weight of SO2F2 in inlet stack gas is 0.0361 g/mole and PP applied the corrected value of 0.0380 g/mole in M_d,in calculation. (In E_SF6,in,y calculation, increasing of M_d,in,y will decrease E_SF6,in,y and when E_SF6,in,y is decided as E_SF6,y, BE will be decreased. In this monitoring period, E_SF6,in,y is decided as E_SF6,y as a result that E_SF6,in,y is the minimum value among E_SF6,in,y, 0.48*C_SF6,y, and 0.48*CSF6,hist. Refer to ER calculation formula (6) ~ (13) in the methodology )
In this monitoring period, the applied mole weight of SO2F2 consists only 0.133% of total mole weight, M_d,in(28.5542 g/mole). Reflecting SO2F2 in M_d,in calculation as a result of detecting SO2F2 in inlet stack gas will decrease BE and also ER will be decreased. From the calculation formula of E_SF6,in,y, if SO2F2 mole weight is not detected in inlet stack gas during this monitoring period, + 0.065% change in E_SF6,in,y would be calculated and it would be resulted to increase BE as same rate under condition that all other monitoring parameters were correctly verified at the time of this devation submission (under this condition, E_SF6,in,y is calculated as 12.692 ton SF6).

We are hereby requesting the approval of the deviation that described above for the period of 21 October 2010 to 28 February 2011.

>>
This proposed deviation from the monitoring plan does not require an amendment to the approved monitoring plan used by the project activity, as the deviation is only for a specific period up to the date that stock change of purchased SF6 is not verifiable with clear evidences or SO2F2 are detected as over 100ppmv gas in M_d,in.
Now PP has measured and reorded the stock amount of SF6 cylinder at the site(gas room) regularly since 1 March 2011 (the starting date of 2nd monitoring period ).
And for M_d,in , it can be identified in next annual test whether the SO2F2 is detected as over 100ppmv in the inlet stack gas or not. The 2nd annual test was implemented on 2 December 2011 and the result shows that SO2F2 is detected as over 100ppmv in the inlet stack gas. If this situation is considered to be permanent during crediting period, MP revision will be implemented in next periodic verification.

>>
1. Monitoring of the consumption of SF6 (C_SF6,y)
SF6 consumption (C_SF6,y) is used for determining one of major factor of baseline emission. E_SF6,y is defined as SF6 baseline emissions in year y(tonnes SF6) and calculated as E_SF6,y=min{E_SF6,in,y ; 0.48*C_SF6,y ; 0.48*C_SF6,hist}. If C_SF6 that is one of parameters to determine E_SF6 is measured conservatively, the baseline determination will be a conservative approach.
As mentioned in the above section, this deviation will ensure applying of conservative value in SF6 consumption for the amount of emission reductions calculation during the monitoring period and does not cause overaccounting of the amont of real emission reductions.
(refer to ER calculation formula (1) ~ (5) in the methodology).

2. Measuring and calibration method of SO2F2 pertaining to M_d,in
In the project activity, SO2F2 gas is detected as over 100 ppmv in inlet gases. The gas could not be analysed using QMS technically in accordance with AM0078 version 01.1. Thus, impact of the deviation on the emission reduction is substituted by comparison result of SiF4 using QMS and FTIR described in the above section “Desciption of the request for deviation”. In conclusion, it is believed that the deviation in analysis method for SO2F2(from calibration method with QMS to library method with FTIR) does not increase the amount of emission reductions taking into account scientific equivalence in analytical methods and even refelction of uncertainty. And the identification of SO2F2 results slightly in decreasing of Emission reduction,
(refer to ER calculation formula (6) ~ (13) in the methodology)