Setting a hybrid style gives me a "Segmentation fault"

Trying to set different styles of interaction with the following code

pair_style hybrid gran/hooke/history 1e-4 NULL 1e-5 NULL 0.0 1 lj/cut 2.5e-6
pair_coeff * * gran/hooke/history
pair_coeff 1 3 lj/cut 1.0 1.0 2.5e-6

gives me a segmentation fault.

This is a NUFEB simulation, where I am trying to simulate different interactions between cells (atoms type 1) and ganular activated carbon (atoms type 3).

# NUFEB simulation with methanogens and GAC floor

units si                                   	# using si units
atom_style      coccus                      # using nufeb atom style

# map array: find atoms using indices sort 1000 5.0e-6: sort every 1000 steps with 5.0e-6 binsize
atom_modify     map array sort 1000 1e-6

# periodic boundaries in x and y fixed boundary in z
boundary        ff ff ff

# forces between local and ghost atoms are computed in each processor without communication
newton          off

processors      * * *                      # processor grid

# communicate velocities for ghost atoms
comm_modify     vel yes

# guarantee that enough atoms are communicated to correctly compute
comm_modify     cutoff 2e-6


# Shift the lattice grid 0.5 unit so that atom can be created in the center of the grid
lattice sc 1e-6 origin 0.5 0.5 0.5
region reg block 0 20 0 20 4 5

create_atoms 1 random 80 31324 reg

# create the GAC floor - this is influenced by the lattice command
region gac_floor block 0 20 0 20 0 1
create_atoms 3 region gac_floor

region outside_gac_floor block 0 20 0 20 0 1

set type 1 density 150
set type 1 diameter 1.3e-6
set type 1 mass 1.725e-16
set type 1 outer_diameter 1.3e-6
set type 1 outer_density 30

# define attributes for type 3, set gac diameter to 1e-6, make it consistent with lattice size
set type 3 density 150
set type 3 mass 1.725e-16
set type 3 diameter 1e-6
set type 3 outer_diameter 1e-6
set type 3 outer_density 30

# set the groups
group met type 1
group eps type 2
group non-gac type 1 2
group gac type 3

# setting neighbour skin distance and style
neighbor        7e-7 bin

# rebuild neighbour list if any atom had moved more than half the skin distance
neigh_modify    check yes

# select grid style
grid_style      nufeb/chemostat 3 h2 co2 ch4 2e-6

# test grid comm
#run_style       test/comm_grid
#run             0

# set substrates initial concentration
grid_modify     set h2   nn nn nn 10 10
grid_modify     set co2   nn nn nn 10 10
grid_modify     set ch4   nn nn nn 0 0

# define pair styles
pair_style hybrid gran/hooke/history 1e-4 NULL 1e-5 NULL 0.0 1 lj/cut 2.5e-6
pair_coeff * * gran/hooke/history
pair_coeff 1 3 lj/cut 1.0 1.0 2.5e-6

# NVE integration with maximum distance limit
# only update position for non-gac atoms
fix nve non-gac nve/limit 1e-8

# monod reaction fixes
fix monod_met met nufeb/growth/methanogen h2 6e-6 co2 2e-4 ch4 growth 1.92e-4 yield 1.6 decay 0 epsyield 1 epsdens 30

# diffusion reaction fixes
fix diff_sub all nufeb/diffusion_reaction h2 4.5e-9
fix diff_co2 all nufeb/diffusion_reaction co2 1.186111e-9
fix diff_ch4 all nufeb/diffusion_reaction ch4 1.488889e-9

# biological model fixes
fix div met nufeb/division/limited 1.36e-6 30 1234 region_blocked gac_floor
fix eps_ext met nufeb/eps_extract 2 eps 1.3 30 5678

# mechanical model fixes
fix wall all wall/gran hooke/history 1e-3 NULL 1e-4 NULL 0 0 zplane 0.0 8e-5
fix eps_adh all nufeb/adhesion/eps eps 1e-6
fix vis all viscous 1e-5
# fix ID atomsID nufeb/shear shearRate viscosity +x (or -x or +y or -y)
fix s1 non-gac nufeb/shear 0.2 0.001 +x

# pressure computation
compute vol all nufeb/volume
compute ke all ke
variable one equal 1.0
compute press all pressure NULL pair vol v_one
variable press equal "(c_ke + c_press) / (3.0 * c_vol)"
variable mass equal "mass(all)"

variable nmet equal "count(met)"
variable neps equal "count(eps)"
compute mycon all nufeb/ave_conc
variable ch4 equal c_mycon[2]

# file output
shell mkdir vtk_gac_sheet
dump 1 all vtk 10 vtk_gac_sheet/dump*.vtu id type diameter
dump 2 all grid/vtk 10 vtk_gac_sheet/dump_%_*.vti con

# thermo output
thermo_style custom step atoms v_press v_mass v_nmet v_neps c_mycon[*]
thermo 1
thermo_modify lost ignore flush yes

# issue run command
run_style nufeb diffdt 1e-4 difftol 1e-12 pairdt 1e-2 pairtol 1 pairmax 1000 diffmax 50000
timestep 1000
run 150

NUFEB uses an outdated version of LAMMPS, so I cannot update LAMMPS to a newer one. Could just the older version be the reason for this problem? Or should I specify some other parameter that LAMMPS is not finding, just reporting that it can’t find its value with “address not mapped”?

It is impossible to say so from remote. It could also be a bug in the NUFEB code.
Please see 11.4. Debugging crashes — LAMMPS documentation for some methods to narrow down where segfaults happen.

If you want further help, you need to create a complete, small input deck that a) causes the segfault and b) only contains commands that are available in the latest LAMMPS release, and post it here.

An alternative to get assistance would be to post an issue in the NUFEB development repository. This project seems to be under active development, so chances are much better that the developers there know where possible issues could be.