mptr¶

Multiple species ingrowth particle tracking. Note that data for each numbered group must be input. The other input is optional.

• Group 1 - NSPECI, MAXLAYERS, MAX_PARTICLES, RIPFEHM, MAX1D
• Group 2 - POUT, PRNT_RST
  • or when PRNT_RST ≥ 20, selected output parameters
• Group 2 - POUT, PRNT_RST, PRNT_VARNUM ( 1 … 6)
  • Optional keyword “tcurve” is input to indicate that transfer function curves should be input to model matrix diffusion. It is followed by NUMPARAMS and TFILENAME.
    • KEYWORD
    • NUMPARAMS, FFMAX
    • TFILENAME
  • Optional keyword “zptr” designates zones for breakthrough curves will be defined. It is followed by IPZONE and IDZONE.
    • KEYWORD ‘zptr’
    • IPZONE
    • IDZONE(I) I = 1 to IPZONE
• Group 3 - RSEED, RSEED_RELEASE
  • Optional keyword “wtri” is input to indicate a water table rise calculation should be performed. It is followed by WATER_TABLE. For GoldSim the water table rise calculation is controlled by passing a new water table elevation to the code during the simulation and the keyword is not required.
    • KEYWORD ‘wtri’
    • WATER_TABLE
• Group 4 - DAYCS, DAYCF, DAYHF, DAYHS
  • An optional, flexible input structure involving the assignment of transport parameters is implemented in the particle tracking input to allow multiple realization simulations to use different parameters for each realization. The user invokes this option using the keyword “file” before Group 5, followed by the name of the file that the transport parameters reside in. The applicable transport parameters are defined in Group 5 and Group 9.
    • KEYWORD ‘file’
    • PFILENAME
  • The structure of the alternate parameter file is:
    • NINPUTS
    • PTRPARAM(I) I=1 to NINPUTS
    • .
    • .
    • .
  • [a line of parameters is present for each realization]

The method for assigning a given value of the particle tracking parameter (PTRPARAM) to a specific transport parameter, defined in Group 5 or Group 9, is discussed below. There are an arbitrary number of input lines each representing a given realization of parameter values. In a multiple-realization scenario, the code enters the input file for each realization, and for this input, reads down the corresponding number of lines to obtain the parameters for that realization. For example, for realization number 10, the code reads down to the 10th line of data (line 11 in the file) and uses those parameter values.

Once these parameters are read in for a given realization, they must be assigned to specific transport parameters. This is done in the following way in Group 5 or Group 9. If any of the inputs other than TRANSFLAG are negative, the code takes the absolute value of the number and interprets it as the column number from which to assign the transport parameter. For example, if DIFFMFL = -5, then the diffusion coefficient is the fifth input number in the PTRPARAM array. In this way, any of the transport parameter inputs can be assigned through the alternate input file rather than the input line in mptr. It should be noted that for the colloid diversity model, only K_REV need be negative to indicate values should be read from the parameter file, if K_REV is negative then all five parameters are read from the file, otherwise the equation parameters will be read from the mptr macro. This is to accommodate the fact that the SLOPE_KF may have negative values.

Group 5 is used to define models in which identical transport parameters are assumed to apply. Group 5 data are read until a blank line is encountered. The model number ITRC is incremented by 1 each time a line is read. Model parameters defined in Group 5 are assigned to nodes or zones using Group6.

Optional keyword “afm” indicates the Active Fracture Model input for determining fracture spacing should be used. Optional keyword “dfree” is input to indicate that a free water diffusion coefficient and tortuosity will be entered instead of the molecular diffusion coefficient.

KEYWORD ‘afm’

KEYWORD ‘dfre’

• Group 5 - TCLX(ITRC), TCLY(ITRC), TCLZ(ITRC), APERTUR(ITRC), MATRIX_POR(ITRC)
  • or when ‘afm’ is implemented:
  • Group 5 - TCLX(ITRC), TCLY(ITRC), TCLZ(ITRC), APERTUR(ITRC), MATRIX_POR(ITRC), SRESIDUAL(ITRC), GAMMA_AFM(ITRC)
• Group 6 - JA, JB, JC, ITRC (JA, JB, JC - defined on See JA, JB, JC, PROP1, PROP2, …)

The following groups (Group 7 - 12) are repeated for each species.

• Group 7 - ITH_SPECI, TRAK_TYPE, HALF_LIFE, IDAUGHTER, CONFACTOR, NEWCONFACTOR, CONFTIME, GMOL, P_FRACTION, ASTEP, CFRACTION

Optional keyword “size” is input to indicate that the colloid size distribution model option is being used. It is followed by PART_SIZE and PROBSIZE.

KEYWORD ‘size’

PART_SIZE(I), PROBSIZE(I) - an arbitrary numbers of lines of input, terminated by a blank line.

Optional keyword “dive” is input to indicate that the colloid diversity model is being used. It is followed by FLAG_COL_DAUGHTER, optional keyword “file” and the name of the file containing the CDF table or equation data (a description of the format for the file is provided with the second mptr example below), the TPRPFLAG with optional SIMNUM, optional CDF equation parameters (when “file” is not used), and keyword “irreversible” or “reversible” with FLAG_LOG.

Note that optional KEYWORDs “size” and “dive” are only used when colloid transport is enabled.

• Group 8 - LAYERS
• Group 9 - LAYER_I, TRANSFLAG, KD, RD_FRAC, DIFFMFL

or for simulations using “dfree”:

• Group 9 - LAYER_I, TRANSFLAG, KD, RD_FRAC, H2O_DIFF, TORT_DIFF

or for simulations with colloid (TRANSFLAG < 0):

• Group 9 - LAYER_I, TRANSFLAG, KD, RD_FRAC, DIFFMFL, KCOLL, RCOLL, FCOLL

or for simulations with colloid using “dfree”:

• Group 9 - LAYER_I, TRANSFLAG, KD, RD_FRAC, H2O_DIFF, TORT_DIFF, KCOLL, RCOLL, FCOLL
• Group 10 - NS
• Group 11 - JA, JB, JC, TMPCNSK

Note that because the number of source terms is controlled by the value entered for NS, Group 11 input is not terminated with a blank line.

• Group 12 - PINMASS, T1SK, T2SK

For transient source terms, Group 12 is repeated for each time interval and terminated with a blank line. Groups 11 and 12 are repeated for each source term (from 1 to NS).

For decay-ingrowth calculations, when the particle injection period is too small (for example, 1.E-4 days) compared to the half-life of the radionuclides and the half-life is large (for example 1.E+9 days), numerical errors in the decay-ingrowth calculation may arise due to truncation error. To get better accuracy, the user should try to increase the length of the injection period.

For particle tracking simulations using the transfer function method (see See Transfer function curve data input file for input file format), it is sometimes desirable to identify the parameter ranges over which the two- and three-parameter type curves are accessed, so that an assessment can be made regarding the density of transfer function curves in a given part of the parameter space. If the flag output_flag in the transfer function file is set to “out”, the code writes the real*8 array param_density to the *.out file in the following format:

For regular parameter spacings, the output is:

i = 1, nump1
   j = 1, nump2
       k = nump3

           write(iout.*) param_density(i,j,k)

       end do
   end do
end do

For two-parameter models, only the i and j loops are used. The value of param_density is the number of times any particle passes through any node at those values of the parameters. This allows the user to identify regions in which a greater density of transfer functions may be required. For the option ‘free’ in which there is no structure to the parameter grid used for the transfer function curves, nump1 is the total number of curves, and nump2 and nump3 are equal to 1.

If “ptrk” was written into the “.ini” file in the original run, the particle data in the “.ini” file are read and used to initialize the particle tracking simulation (Case 2). In this instance, the number of particles (NPART) must be set the same for the restart run as in the original run or the results will be unpredictable.When restarting a particle tracking simulation, certain input data are overwritten by information in the “.ini” file. These parameters include RSEED, RSEED_RELEASE, PCNSK, T1SK, and T2SK. Other input parameters can be set to different values in the restart run than they were in the original run, but of course care must be taken to avoid physically unrealistic assumptions, such as an abrupt change in transport properties (Group 4 input) part way through a simulation.

A final note on restart calculations is in order. A common technique in FEHM restart calculations is to reset the starting time at the top of the “.ini” file to 0 or in the time macro so that the starting time of the restart simulation is arbitrarily 0, rather than the ending time of the original simulation. This is useful for the example of the steady state flow calculation, followed by a restart solute transport calculation. Although this technique is acceptable for particle tracking runs that are initiated only upon restart (Case 1), it is invalid when a particle tracking run is being resumed (Case 2). The reason is that all particle times read from the “.ini” file are based on the starting time of the original simulation during which the particle tracking simulation was initiated.

The following is an example of mptr. A multiple-species decay-chain (\(\rightarrow 2 \rightarrow 3\)) is simulated, with decay half lives of the species equaling 10,000, 3,000, 10,000, and 4,000 years, respectively. In this simulation a maximum of 3 property layers are specified although only 1 layer is used, the maximum number of particles is specified to be 1100100, and FEHM is run in stand-alone mode. Concentrations will be computed as number of particles per unit fluid volume and no output will be written to the “.fin” file. Use of the ‘zptr’ keyword indicates that a single zone will be defined for breakthrough curve output which will be written to the “.out” file. The random number seed is defined to be 244562. The particle tracking solution is enabled at 0.1 days, and disabled at 3.65e8 days, while the flow solution is disabled at 38 days and re-enabled at 3.65e8 days. Dispersivity in the X-, Y-, and Z- directions are defined to be 0.005 m, the mean fracture aperture is 0.0001 m, and the matrix porosity is 0.3. Particles for species 1 are injected at a constant rate from 0 to 5,000 years, and species 2, 3, and 4 are formed through the decay reactions, with no input at the inlet. Advection and dispersion (without matrix diffusion) is being modeled. The retardation factors for the four species are 1, 1, 1.9, and 1, respectively (i.e. only species 3 sorbs).

In the second example, transfer function data is used along with the active fracture and colloid diversity models. The the cumulative probability distribution function (CDF) retardation parameters for the colloid diversity model are entered in an external file using the table format. The format for the new input files associated with the colloid diversity model are:

For ptrk/sptr/mptr simulations with TPRP_FLAG = 11 or 12:

• Header line indicating the species number (always 1 for ptrk/sptr simulations)
• Multiple tables, each with the following format:
  • One line specifying the realization number
  • Multiple lines with two columns of data (real) representing rcdiv and probdiv. Note that probdiv should start at 0 and end with 1.
  • A blank line is used to specify the end of the distribution table.
• A blank line is used to specify the end of the file

For mptr, a distribution will need to be entered for each colloid species. Therefore, the header line and tables are repeated for each colloid species.

For ptrk/mptr simulations with TPRP_FLAG = 13 or 14:

• Header line containing comments
• Multiple lines, each line containing realization number, \(b, m, k_r, R_{min}, R_{max}\)
• A blank line is used to specify the end of the file

For sptr simulations with TPRP_FLAG = 13 or 14:

• Header line containing comments
• Multiple lines, each line containing realization number, \(b, m, k_r, R_{min}, R_{max}, \alpha_L\)
• A blank line is used to specify the end of the file

Particle statistics data for the cumulative number of particles that have left the sytem and the number of particles that left during the current timestep are output.

With the file rcoll_data.dat as: