Pagina 83 di 89 Transfeu WP2 Subtask 2.1.2 Development of small-scale test method for fire effluents Step 2: Influence of 2 nd filter on FTIR sampling line to prevent HCl and other halogen gases trapping Plans according to DoW: The necessity to safeguard the integrity of the sensitive FTIR cell requires the use of a system of filters that remove the particles and remains efficient for the whole duration of the test. LSFIRE, LNE, SP and CUR will study the following parameters: type of material of the filters, porosity, and level of temperature. Work done and results 1. Forward According to the plans of Transfeu DoW WP2 subtask 2.1.2 step 2, LSFire, have planned and carried out some tests to verify the best configuration of filtering system in FTIR sampling line for removing the smoke particles in order to safeguard the integrity of the sensitive gas cell and so keeping it efficient for the whole duration of test. In particular, additional tests were performed to verify the influence of 2 nd filter (placed in in-let gas cell point) to prevent HCl and other halogen gases trapping. For some gas species it was observed, rather during the calibration step by standard mixture bottles than carried out tests using reference liquids in order to proceed on a preliminary system verification (burning as a single component and as a defined mixture of all of them), differences due to factors which were subjected to as additional study. In particular, HCl (and probably HBr also) seems to be partially trapped in the part of sampling line between the smoke chamber and the FTIR gas cell. It was supposed that the presence of cold points and the effect of carbon soot on filters surfaces may be a possible causes of halogen gas condensation. After a deep discussion within Transfeu WP2 meeting and successively at Scientific panel, it was decided to verify the influence of filter units performing some tests using directly standard gas mixture bottles (the same used for calibration) as a defined sampling source, and sampling line described in the last approved test protocol in different configurations. The experimental part has been conducted in LSFire laboratory of Controguerra (Italy) in March 2010. 2. Testing set up The general configuration (probe, sampling line, cell, cooler, etc.) was set according the last test protocol defined in preliminary steps (doc.. Transfeu WP 2.1.3 Small-scale test method for fire effluents -DRAFT 2). It was also decided to use 2 different concentrations of HCl (standard certificated gas mixture bottles) in nitrogen: 200ppm and 500ppm. Three different configurations of FTIR sampling system were set for study purpose: Configuration 1
Pagina 84 di 89 Flushing the HCl gas mixtures from the bottle to the internal probe of chamber and keeping the only main filter (placed in oul let of probe) heated at 180 C, sampling line (180 C), gas cell (180 C) and completely removing the 2 nd filter housing. Configuration 2 Flushing the HCl gas mixtures from the bottle to the internal probe of chamber, the main filter (placed in oul let of probe) heated at 180 C, sampling line (180 C), the 2 nd filter heated at 130 C and gas cell (180 C). Configuration 3 Flushing the HCl gas mixtures from the bottle to the internal probe of chamber, the main filter (placed in oul let of probe) heated at 180 C, sampling line (180 C), the 2 nd filter heated at 170 C and gas cell (180 C). In all 3 configurations (cone radiator heater off) the flushing of gas mixture was monitored in continuous way by absorbance spectra collection (concentration vs time) and by cell pressure data acquisition (cell pressure vs. Time) for concentration data correction. The schemes and procedure adopted to carried out the tests are following reported: Configuration nr.1 (no 2 nd filter) 1st filter (180 C) Sampl. line (180 C) P No 2nd filter Gas cell (180 C) Probe F4 Smoke chamber V3 Cooler (4-20 C) V4 O4 F2 V1 V2 O2 Nitrogen bottle HCl + Nitr bottle Configuration nr.2 (with 2 nd filter heated at 130 C)
Pagina 85 di 89 1st filter (180 C) Sampl. line (180 C) P 2nd filter heated at 130 C Gas cell (180 C) Probe F4 Smoke chamber Cooler (4-20 C) V4 O4 V3 F2 V1 V2 O2 Nitrogen bottle HCl + Nitr bottle Configuration nr.3 (with 2 nd filter heated at 170 C) 1st filter (180 C) Sampl. line (180 C) P 2nd filter heated at 170 C Gas cell (180 C) Probe F4 Smoke chamber V3 Cooler (4-20 C) V4 O4 F2 V1 V2 O2 Nitrogen bottle HCl + Nitr bottle
Pagina 86 di 89 Index: PTFE pipes; O2 = rotameter; O4 = rotameter; V1, V2, V3 and V4 = valves General procedure a. Wait that all parts in the scheme reach the temperature steady-state for a minimum time of 30 min. b. Keeping close the standard gas mixture and nitrogen bottles, open totally valves V2, V3 and close valve V3. c. Open nitrogen bottle and adjust F2 to 2 l/min. d. Open V3 slowly until atmospheric pressure looking P output monitoring. e. Adjust F4 with V2 to 1.5 ± 0.1 l/min. f. Wait 1 minute for flow rate and pressure (P) stabilization g. Collect background spectrum and measure the pressure P during collection h. Start the continuous collection of spectra i. At 25s from the starting point close nitrogen bottle j. Open standard gas mixture bottle (HCl) and adjust F2 to 2 l/min. k. Adjust V3 slowly until atmospheric pressure looking P output monitoring. l. Adjust F4 with V2 to 1.5 ± 0.1 l/min. m. Continue in continuous absorbance spectra collection and cell pressure data for at least 10min. n. Close the standard gas mixture bottles and purge all the line introducing nitrogen for a minimum period of 5min. 3. Results The comparison between the different configurations and 2 standard concentration mixture bottles in terms of HCl trend is reported in following graph.
Pagina 87 di 89 The results shown that no differences occur in terms of HCl decreasing between the configurations which not include and include the 2 nd filter, and in the same way no differences are detected if the 2 nd filter is heated at 130 C or 170 C. Currently, it is clear that if more or less relevant part of HCl was trapped along the sampling line, the phenomenon is not caused by its condensation into cold points. If the flushing flow from the bottle do not have any losses, the HCl concentration in short time reaches the correct value without any problems coming from the sampling line even if pass though 2 filtering systems. The HCl losses may be caused, during testing on materials, by a carbon soot accumulation on filtering membranes which become a further active filter where a part of halogen gas will be absorbed and trapped. Considering that this process is strictly depending by the amount of soot produced from the different test samples, it is impossible to prevent their effects in any of bench scale test as well as for intermediate, full and real scale methods. 4. Additional comments and recommendations During this work, some additional technical aspects were came out and must be submitted to the people who will be involved in experimental work. The maintenance of standard mixture bottles/cylinders and the correct set-up/use of sampling configuration during the calibration step are extremely important to avoid relevant mistakes in concentration prediction, especially when halogen gas species mixture cylinders are involved. In fact, before the use of these bottles is important that they must be stored for a minimum of 24 h at 20-25 C of ambient temperature and then should be rolled for at least 10 min before collecting to the calibration line.
Pagina 88 di 89 It shall be done to unsure that stratification and condensation in bottle phenomena will not occur and cause a wrong concentration flow through the calibration line (lower concentration than the real). A second main factor to be taken into account is the use of T line system in calibration line (valve V2 and rotameter O2 in configuration schemes) that was introduced and described in FTIR analyser calibration procedure (par. 7.4.2 and 7.4.3 of doc.. Transfeu WP 2.1.3 Small-scale test method for fire effluents - DRAFT 2). The reason is to avoid the directly use of pressure reducer of gas bottles to reach a low flow rate as 1,5 l/min. In fact, it may cause a mistake in concentration flow though the line because the valve of pressure reducer works on a too low pressure and cause a partial and uncorrected/unstable free flushing from the cylinder to the sampling line. Examples are reported below.
Pagina 89 di 89 Pictures Claudio Baiocchi / Silvio Messa (LSFire)