ML20203G093

From kanterella
Jump to navigation Jump to search
Analysis of Physical & Chemical Properties of Original & Replacement Activated Carbon
ML20203G093
Person / Time
Site: Perry  FirstEnergy icon.png
Issue date: 07/29/1986
From:
CLEVELAND ELECTRIC ILLUMINATING CO.
To:
Shared Package
ML20203G077 List:
References
NUDOCS 8607310321
Download: ML20203G093 (12)
v • d • e


Text

m.msa-me_.#, y a. ms m wa J - . -

--4 Wie h

'4 E M.asa_. .J __~J Ja4 .. mag.4_ g....s__.A_4y. ,, , #. . ._.a.w>J.4 -.4 _

,2 __A aA-

[

4 O

f f

i 4

1 a

2 i

i 5 &

n i

l. ANALYSIS OF THE PHYSICAL AND CHEMICAL PROPERTIES OF THE l

i

ORIGINAL AND REPLACEMENT ACTIVATED CARBON 1

I i

+

4 l

i 8607310321 860729 PDR ADOCK 05000440 S PDR O

e O

ANAINSIS OF THE PHYSICAL AND CHEMICAL PROPERTIES OF THE ORIGINAL AND REPLACEMENT ACTIVATED CARBON General physical properties of activated coconut shell carbon, manufactured for use in noble gas delay systems, are explained in this section as they relate to ignition tenperature under the dynamic conditions experienced at Perry. Testing results are provided for the original in situ carbon, the unused Barneby-Cheney Co. carbon from on-site warehouse supplies and the NUCON

' supplied replacement carbon.

General P aperties of Noble Gas Delay Carbon Activated carbon manufactured for service in noble gas delay systems (off-gas b

v treatment) is typically virgin unimpregnated coconut shell base carbon, activated from char at high temperature (approximately 1300'C) in the presence of a mixture of steam and oxygen, to produce a microporous structure suitable for physical adsorption of noble gases, such as Xenon and Krypton. Carbon Tetrachloride " activity" testing (ANSI / ASTM D3467-1976) is performed to verify the adequaev of the carbon to adsorb. Values in the 30% to 35% range indicate the presence of the correct type and quantity of pores which will effectively l

adsorb relatively small noble gas molecules. Prior to its acceptance for i

service and installation into system adsorber vessels, the carbon is dried to less than 2 percent (by weight) moisture and tested for its sorptive ability to " delay" Xenon and Krypton under either static or dynamic conditions. Other

, minimum phyJ;ical requirements must also be met prior to acceptance, such as ignition temperature, hardness and density; however, adsorbed moisture is the l

most important of these requirements for ensuring optimum performance.

l I _ . _ . . _ . . _ _ . . - _ .

i O

V Moisture is the major adsorption inhibitor since water vapor molecules effectively occupy adsorption sites in the carbon micropore structure, making

~

these sites unavailable for noble gas adsorption.

Ignition Temperature Considerations of Activated Carbon

'Ihe ignition temperature of activated carbon is determined by a standard method, ASTM D 3466-1,83, under controlled conditions. This ignition temperature cannot be construed as the probable ignition temperature of the same carbon sample under conditions other than the conditions specified by Asm D 3466-1983. As specified in ASTM D 3466, a test bed of activated carbon is exposed to an air stream with a face velocity of approximately 100 feet per minute. This air stream is heated in a programmed manner to produce a temperature rise of 10'C per minute. When the carbon temperature is within 50*C of the anticipated ignition temperature, (or at 150*C if this temperature is unknown), the programmed rise is reduced to 2-3'C per minute. The temperature noted just prior to a sudden, sharp rise, as indicated on a strip chart, is the defined ignition temperature.

The ignition temperature of activated carbon can be increased or decreased by

, a number of factors. Gas velocity is critical if two (2) results are to be l

compared. At lower velocities, removal of heat is reduced and lower ignition temperature results are observed. In the case of the adsorber beds at Perry, l

the face velocity across the adsorber beds which burned was approximately 4 fpm (50 scfm per train divided by 12.6 square feet cross sectional area).

For this uignificant difference in face velocity with respect to the standard I

AS E test conditions, it can be assumed that the ignition temperature of the

i.

(D v activated carbon in tbc adsorber vessels, at the actual velocity during the time of combustion, was much lower than its certified AS M value. Results of analyses performed at Nuclear Consulting Services, Inc. (NUCON), on samples from the adsorber vessels, indicate ignition takes place at a temperature as low as 153*C with a face velocity of 4 fpm.

Other variables can and will have significant effects on the temperature at which activated carbon will ignite and burn. Moisture content, rate of heat application and adsorbed contaminants are significant contributors which can raise or lower ignition temperatures. Desorption of moisture from carbon, (an endothermic process), can effectively raise the ignition temperature by carrying away heat; however, it is necessary for significant amounts of adsorbed moisture to be rapidly desorbed for this effect to be noticeable.

O Test results revealed that no more than 2% moisture was present in any of the V

Perry samples. This is an insufficient amount of moisture to raise ignition temperatures detectably. The rate of heat application is also critical in igniting activated carbon. A high heat application rate, which would inhibit heat removal during oxidation prior to ignition, (an exothermic process), will result in a lower temperature of ignition. Adsorbed contaminants in sufficient quantities, in the form of volatile organics, can cause ignition of carbon at greatly lowered temperatures if desorbed volatiles approach flammable concentrations and ignite, causing rapid oxidation and ignition of O

O

- . . ~_ -. ..

the carbon as well. Low boiling organics, such as ketones, alconols and cther solvents, are typical substances which are flammable and easily desorbed to cause ignition of carbon at lowered temperatures. Insufficient quantities of volatiles were found in any of the samples to be considered as a contributor to ignition.

To summarize, specific physical conditions, such as system process flow rate, rate of heat application, moisture content of the activated carbon and presence of adsorbed contaminants could have affected the ignition temperature of the activated carbon in the adsorber vessels at Perry. Specific analyses were performed to determine if any of these conditions were present in sufficient quantities to affect the ignition temperature of the activated carbon in the Perry adsorber vessels. Circumstances of the testing in progress at the time of ignition and the results of analyses performed were evaluated to formulate a most probable scenario for ignition and subsequent combustion. Results of these analyses are presented in the attachments and are evaluated under Tab 5 " Analysis and determination of the mode of ignition for both combustion events".

O

Samples and Analyses

~

Following combustion in adsorber bed vessels 14A and 14B, a sampling program was designed and a temporary instruction, RTI-0008, (Attachment 1), was written to provide sampling methodology as well as rationale for each type of sample taken. This instruction and the accompanying analyses were designed to provide data for the determination of the most probable source of ignition and certification of the adsorbent for further service in accordance with General Electric Specification 21A9375, (Attachment 2).

The following samples were taken in accordance with RTI-0008:

1. Vessels 14 A&B, top: Samples were removed from the top of these vessels to investigate the extent of combustion, the cause of ignition and the suitability of the adsorbent for further service.
2. Vessels 12 A&B, lower thermocouple ports: Samples were removed from the lower thermocouple ports of these vessels to investigate the presence of possible organic contamination and to provide a basis for comparison of installed carbon to carbon which had not experienced combustion.
3. Vessel 14B, lower thermocouple port: Samples were removed from the lower thermocouple port of this vessel to investigate the effects of combustion on adsorbent material downstream of the event.

O

4. Barneby-Cheney Type 483, Lot 1605 (from warehouse): Samples were removed from warehouse stock for comparison between installed adsorbent and the unused supply. -

Activated carbon analyses were performed by Nuclear Consulting Services, Inc.

(NUCON), using current ASM or NUCON approved procedures.

1. The following analyses were performed to ensure suitability for service, based on General Electric Specification 21A9375:
a. Percent Moisture, ASE D 2867
b. Apparent Density, ASW D 2854 O
c. Hardness Percent, ASTM D 3802
d. Particle Size Distribution, ASTM D 2862 i
e. Ignition Temperature, ASTM D 3466 i
f. Dynamic Adsorption Coefficient, Kr, NUCON PROCEDURE 30
g. Dynamic Adsorption Coefficient, Xe, NUCON PROCEDURE 30 l

l O

0

2. %e following analyses were perforned to provide data to determine the most probable ignition scenario in adsorber beds 14A and 14B:
a. Ignition Temperature, ASIM D 3466 (performed at 4 fpm face velocity)
b. Volatile Weight Percent, NUCON PROCEDURE 121
c. Ash Percent, ASTM D 2866
d. Particle Size Distribution, ASTM D 2862

, Test Results for Barneby-Cheney Co. Carbon A tabulation of test results for the Barneby-Cheney Co. carbon, sairpled following the first combustion event, is provided in Attachment 3. Tests of the carbon were performed June 30, 1986 thru July 6, 1986 by Nuclear Consulting Services, Inc. The test results submittal from NUCON is included in Attachment 4: NUCON 06C0869/01.

The ignition test analyses were performed to confirm or discount factors which may have contributed to combustion in the adsorber vessels. Factors investigated which reduce ignition temperature were: reduced particle size (presence of " fines"), presence of adsorbed volatile matter, adsorbed moisture content, ash content and low air velocity during the testing.

O Particle size distribution analysis of the carbon sampled from both the V

adsorber trains and compared to the warehouse supply revealed that no significant size degradation had occurred as a result of combustion.J Adsorbent sampled from beds not involved in combustion had a particle size distribution virtually identical to that of the unused warehouse supply. It was concluded from these results that the presence of carbon fines was not a contributing factor to lowering the effective ignition' temperature of the adsorbent.

Adsorbed volatile matter was not detected on any adsorbent samples in quantities sufficient to lower the ignition temperature of the carbon in the system. Moisture content of all samples removed from the vessels was extremely low (all samples contained less than 1% by weight).

A V

Ash content was found to be in the normal range for activated carbon. For all samples analyzed, the results were less than 2% by weight, ash. Therefore, the presence of excessive ash was not a contributor to ignition.

The significant analysis result of all the tests performed was the resultant Asm ignition temperature at 4 fpm face velocity. It was confirmed that the adsorbent in the vessels could ignite at temperatures as low as 153'C, a much lower temperature than the minimum specified 250*C, which is determined at 100 fpm face velocity.

O o l

l

1 1

l i

/3 The activated carbon removed from the adsorber vessels, as well as that which L) was sampled from the unused warehouse supply, was shown to meet the minimum requirements of apparent density, hardness, moisture content, particl_e size distribution and ignition temperature in accordance with General Electric Specification 21A9375.

Results of adsorption coefficient determinations for Krypton and Xenon were found to be acceptable for service in noble gas delay systems such as off-gas treatment. NUCON procedure number 30 was used in place of the G.E. procedure.

The NUCON procedure number 30 provides a more conservative test method by specifying a larger, more representative sample size, as well as a more accurate replication of actual system operating conditions during the test.

Correlation of results is made by comparison of test results of carbon from the adsorber vessels to unused carbon from the warehouse, which has already been certified in accordance with the G.E. specification.

The low process flow velocities coupled with sustained heat application appear to be significant contributors to the initial carbon ignition. This process will be expanded upon in the section of this report dealing with determination of the most probable mode of ignition and combustion propagation.

Test Results for NUCON Replacement Carbon A tabulation of test results for the NUCON supplied replacement carbon is provided in Attachment 5. Testing for chemical and physical properties of the NUCON carbon have been performed which are identical to or more conservative than those tests performed on the original Barneby-Cheney Co. supplied carbon.

O

.- - _ _ _ -~ _._ _

4 i

1 f All NUCON supplied carbon meets or exceeds requirements for activated carbon i for use in the Off-Gas System adsorber beds. General Electric has also approved the use of NUCON activated carbon in the Perry Off-Gas System.

)

4 l

I i

l I

i s

4 O

I l

l

, i i

+

e O

10 -

O ATTACHMENT 1 Radiation Temporary Instruction - (RTI) - 0008 Off-Gas System Activated Carbon Analysis O

I O

_. . . _. . _ _ - . _ - _ _ - _ _ _ - _ - _ _ _ _

Retrieved from "https://kanterella.com/w/index.php?title=ML20203G093&oldid=3067878"