The Master Chemical Mechanism
The EUPHORE facility consists of two FEP foil hemispherical chambers each with approximate volume of 200 m3, and high transmission of both visible and UV light through the chamber walls. These large photochemical smog reactors are equipped with complete analytical instrumentation systems for monitoring concentrations of reagents and products, and are ideal for studying chemical processes in conditions close to those found in the ambient atmosphere.
When simulating EUPHORE chamber experiments using MCM chemical
schemes it is important to include an
HONO concentration, and an
mechanism to account for chamber wall reactions, and to take into account
the effect of the chamber on calculated
of stable species should also be accounted for. Some notes on the simulation of
EUPHORE experiments are given in this page, but please note that further
characterisation experiments are underway, and the auxiliary mechanism will be
validated against additional datasets, including some with measurements of
!!! EUPHORE USERS SHOULD WATCH THIS SPACE FOR DEVELOPMENTS !!!
Notes on the comparison of modelled and measured concentrations are also given, and an Excel template file is provided to assist such comparisons.
MCM is a very detailed chemical oxidation mechanism, and to facilitate mechanistic and kinetic interpretation of chamber data, users may wish to reduce the full mechanisms. Objective methods, based on sensitivity analysis, are available (REF). These allow the identification of redundant species and reactions which can be removed without causing unacceptable discrepancies in the important features to be modelled. Following the removal of these reactions and species the quasi-steady-state-approximation (QSSA) can be used to express the concentrations of fast intermediate species algebraically in terms of the other species. The QSSA species are identified by calculation of the error induced by application of the QSSA. The postprocessor, KINALC, can extract the important pieces of information from the sensitivity results output from simulation programs. It can be used to identify redundant reactions and species, and can estimate the instantaneous error of QSSA species and thus enables the proper selection of QSSA species.
One of the most controversial issues of chemical mechanism development and evaluation with smog-chamber data is the influence of chamber-dependent reactions on the reactivity of the chemical system under observation. There are three major groups of processes that contribute to chamber effects :
The simple auxiliary mechanism presented below has been constructed using EUPHORE characterisation experiments and full details of the development are given in a forthcoming publication (REF). WHNO3 and WO3 represent wall bound HNO3 and O3 respectively. All other chamber related reactions were found to have negligible effects on simulations of ethene photosmog experiments.
% 0.7D-5 : NO2 = HONO ;
% 1.6D-5 : NO2 = WHNO3 ;
% 3D-6 : O3 = WO3 ;
Another important parameter for correctly simulating chamber experiments is initial HONO concentration, which particularly effects the timing of simulations in the early stages of the photo-oxidation. The initial HONO concentration is dependent on the intial NOx. Based on simulations of ethene experiments, where the chemical mechanism is well understood, appropriate initial HONO values are estimated to be :
In general MCM uses photolysis rates calculated from parameterisations as a function of solar zenith angle. These derive from a two stream isotropic scattering model for clear sky conditions. Where measured photolysis rates exist these should be used in place of the calculated values.
For experiments carried out in EUPHORE chamber A, usually J(NO2) is the only measured photolysis rate available. In this case measured J(NO2) can be used in the model (J<4> in MCM) and for all other photolysis processes calculated rates can be modified with scaling factors to account for the transmission of the chamber walls, backscattering from the chamber floor and cloud cover. The variations from day to day and during the experiments, resulting from cloud cover, can be accounted for by considering the difference between measured and calculated J(NO2) at any given time during the experiments. This variable scaling factor, f1, is applied to all photolysis rates in addition to a factor, f2, designed to account for wall and cloud transmission effects which will be different for different species dependent on their absorption spectrum. A chamber dataset (22/10/97) with the photolysis rates for O3->O1D, HCHO, HONO available in addition to J(NO2), was used to derive scaling factors, f2, based on the deviation between their measured and calculated photolysis rates, normalised by the deviation seen for NO2. For all other photolysis rates the average value of these factors is used for f2.
J<4> = measured J(NO2);
f1 = J<4>/(1.165D-02*(COSX@(0.244))*EXP(-0.267*SECX));
O3->O1D, f2 = 0.85,
HONO, f2 = 0.84,
HCHO->HCO + H, f2 = 1.03,
HCHO->CO + H2, f2 = 1.03,
average f2 = 0.95.
During the experiments a certain amount of air from the chamber is lost through small leaks and withdrawal of air samples for analysis. Clean air is added to compensate for this and some dilution of the reactants and products occurs as a result. Generally, to measure the dilution rate SF6 is added to the reaction mixture as an inert tracer in each experiment and its concentration is monitored by FTIR. The average calculated loss rate of SF6 over the course of each experiment is used as the dilution rate applied to all stable species in the simulations.
MCM is designed to be used with the FACSIMILE language. Model files should contain output instructions appropriate for the application. Use of the PSTREAM command and instruction produces tabulated numerical output of a fixed format and is useful for creating files of model output to compare to measured data. Including the following lines of code in the model file will create a standard output file of reactant, NOx, HOx, O3, and product concentrations as a function of time.
TIME = 36000 + 300*60 % CALL OUTPUT ;
This calls the OUTPUT routine at the start of the experiment (set to 10 a.m. in this example) and every 5 minutes (300 s) thereafter, for a total run of 5 hours (i.e. 60 times). The user should adjust the start time, time resolution, and number of output points as required
PSTREAM 1 7 20 ;
TIME REACT O3 NO NO2 OH HO2 RO2 H2O2 HONO HNO3 NO3
PROD1 PROD2 PROD3 PROD4 PROD5 PROD6 PROD7 PROD8;
This command defines the concentrations to be printed, up to 20 in any one such instruction, and allocates the unit number for the output. The user should modify this line to include the MCM names for the reactant and products of interest.
NOTE : if using FACSIMILE for Windows, SETPSTREAM should be used in place of PSTREAM in this command.
EXEC OPEN 7 "output1.dat" ;
This assigns the output file to the appropriate unit number.
NOTE : if using FACSIMILE for Windows, EXECUTE should be used in place of EXEC in this instruction.
COMPILE OUTPUT ;
PSTREAM 1 ;
This table of numerical values can then be opened in programs such as Excel for further analysis. A template Excel file is provided. The user can paste the data from "output1.dat" into the "model" sheet, paste the measured data into the relevant columns of the "measured" sheet, and plots showing model-measurement comparisons will be generated in the "plot" sheets.