Hydrogen & CNG Tank Capacity and Fast-Fill Heating

Calculate real-gas mass, compressibility, usable energy and range. Simulate gas and wall heating during hydrogen or methane fills with pre-cooling and pressure cutoff.

#CNG #CoolProp #fast filling #heat transfer #hydrogen #real gas #real-gas EOS #SAE J2601

A hydrogen tank capacity calculator needs more than the ideal gas law at vehicle storage pressures. This tool uses the CoolProp reference equations of state to calculate gas density, stored mass and compressibility at your tank's actual pressure and temperature. It is intended for engineers comparing storage concepts, students studying open-system thermodynamics, and vehicle designers making early fuel-budget estimates.

Start with the internal water volume, gas, nominal full pressure and capacity temperature. Use 350 or 700 bar for the usual hydrogen comparisons. For a CNG cylinder capacity calculator, select the pure-methane proxy and enter the relevant cylinder pressure, commonly 200 or 250 bar. Every pressure here is absolute. Subtracting the gas remaining at the minimum usable pressure gives equilibrium usable fuel. Chemical energy uses lower heating value; range divides usable mass by your own fuel consumption. Neither result includes a powertrain efficiency assumption.

The density chart shows why 700 bar hydrogen mass cannot be calculated reliably with ideal-gas density. The hydrogen compressibility factor is greater than one in this region, so the ideal law overstates capacity. The pressure comparison table evaluates the selected gas at one common temperature and volume; its rows are thermodynamic comparisons, not interchangeable vessel ratings.

For fast filling, enter the initial tank state, station supply pressure and pre-cooled supply temperature, planned duration, wall heat capacity and heat transfer coefficients. Constant inflow normally delivers the nominal mass defined at 15 °C. Under Fill boundary · advanced you can prescribe a mass increment and specify temperature at the tank inlet after throttling. The calculation stops at the planned mass or the chosen pressure cutoff, whichever comes first. A cutoff ratio is a study input and does not authorize that pressure for a real cylinder.

The gas energy balance integrates incoming enthalpy, including the flow work that heats a filling vessel. With temperature specified after the valve, inlet enthalpy changes with tank pressure. The wall absorbs heat and loses heat to ambient air. Pressure and temperature therefore evolve together through the real-gas equation of state. Play the filling sequence to see the pressure gauge rise and the uniform gas colour warm. The comparison chart repeats the simulation with supply gas at ambient temperature and with twice the planned duration, keeping the other model parameters fixed. Two published 145 L hydrogen CFD fill cases check temperature and pressure with heat transfer active. A conservative work screen rejects excessively stiff combinations before integration and explains what to change.

State of charge is gas density divided by the density at nominal pressure and 15 °C, following the convention discussed with SAE J2601. The 85 °C hydrogen bulk-temperature flag provides context for hydrogen refueling temperature rise. This is a thermodynamic study, not an implementation of J2601 dispenser tables, pressure ramps or communications.

The lumped wall cannot resolve radial conduction, liner hot spots or inlet-jet temperature gradients. Effective wall capacity and transfer coefficients need calibration for the actual vessel and fill schedule. Pure methane does not represent every natural-gas composition. The tool excludes cylinder strength, fatigue, relief devices, leakage, station cascade dynamics and installation design. Real fueling requires the tank manufacturer's approved protocol and certified dispenser controls; check against the standard in force.

How it is calculated

Capacity uses m=Vρ(p,T)m=V\rho(p,T) and Z=p/(ρRsT)Z=p/(\rho R_sT) from CoolProp 8.0.0: normal hydrogen uses Leachman et al. (2009), methane uses Setzmann–Wagner (1991). The ideal comparison uses the same molar mass and universal gas constant. Equilibrium reserve mass is evaluated at minimum pressure and the capacity temperature. Lower-heating-value energy is usable mass × LHV; range uses the entered mass consumption per 100 km.

The fixed-volume open-system model integrates gas total internal energy UgU_g, uniform wall temperature TwT_w, signed heat to ambient QaQ_a, and accumulated inlet enthalpy EinE_{in}:

m˙=m˙in,U˙g=m˙inhin−hiA(Tg−Tw),\dot m=\dot m_{in},\quad \dot U_g=\dot m_{in}h_{in}-h_iA(T_g-T_w),
CwT˙w=hiA(Tg−Tw)−hoA(Tw−Ta),Q˙a=hoA(Tw−Ta).C_w\dot T_w=h_iA(T_g-T_w)-h_oA(T_w-T_a),\quad \dot Q_a=h_oA(T_w-T_a).

Mass follows m=m0+m˙intm=m_0+\dot m_{in}t exactly. Each stage recovers TgT_g from CoolProp at density m/Vm/V and specific internal energy Ug/mU_g/m, then pressure from (Tg,ρ)(T_g,\rho). For a temperature specified at the station, inlet enthalpy is evaluated at the supply pressure before an ideal isenthalpic restriction and remains constant. For a temperature specified at the tank inlet after throttling, each stage evaluates hin=h(Tin,pg)h_{in}=h(T_{in},p_g) at the current tank pressure. In both modes, E˙in=m˙hin\dot E_{in}=\dot m h_{in} is integrated alongside the gas and wall states; temperature is never assumed to stay unchanged across a throttle. The first-law form follows Yang (2009); adding the wall node conserves gas-plus-wall energy.

A cylinder with hemispherical ends and the entered length/diameter ratio determines AA. Fourth-order Runge–Kutta uses a maximum step of planned duration / integration steps, shortened to 0.35 times the gas or wall thermal relaxation time and to a maximum 10% mass increase per substep. A pressure crossing is located by 24 bisections, rerunning the fractional RK4 step. By default the planned mass is the nominal 15 °C mass; the advanced mass-increment option instead prescribes a chosen mass addition over the same duration. The fill ends at the planned mass or the chosen pressure cutoff. SOC is 100m/[Vρ(pnom,288.15 K)]100m/[V\rho(p_{nom},288.15\,K)]. Cooldown pressure is a separate sealed-tank EOS evaluation at ambient temperature; it is not a simulated cooldown transient.

Woodfield et al. (2008), attached pp. 1 and 6, supplies the illustrative 500 W/m²·K inner and 4.5 W/m²·K outer coefficients. Its experiments use distributed wall conduction; they supply coefficients and qualitative context. Quantitative nonzero-heat-transfer checks use Zhao et al. (2022), not a reconstruction of the incompletely specified Woodfield experiments.

Two published hydrogen CFD cases from Zhao et al. (2022), Tables 1, 3 and 4 and §4.3.3 (attached pp. 4–5 and 10–13), use a 145 L Type III tank, 5 MPa and 273 K initially, 253 K at the tank inlet, and 180 s fills at 8 and 16 g/s. The reported endpoints are 294.4 K / 19.36 MPa and 306.6 K / 37.66 MPa. Both keep gas–wall and wall–ambient heat transfer active. Wall capacity is the published conservative lower bound, 77 kg × 902 J/kg·K = 69.454 kJ/K; the paper's two material heat capacities bracket it up to 70.84 kJ/K. The 500 W/m²·K inner coefficient is independently taken from Woodfield, without fitting to these endpoints; Zhao supplies 6 W/m²·K outside. Reference tolerances are 3 K and 0.6 MPa.

Assumptions

Rigid tank; normal hydrogen or pure methane; one homogeneous gas state and one effective wall temperature, initially equal. Constant station pressure, specified boundary temperature, mass inflow, wall heat capacity and transfer coefficients; no outflow, leaks, reactions or kinetic-energy storage. Station gas is throttled without added line heat. A temperature controlled at the tank inlet implies that the required upstream enthalpy changes with tank pressure. Inner and outer areas are both approximated by the internal vessel area. The 80 kJ/K wall default is an illustrative effective capacity, not a material catalogue value.

LHV values are rounded engineering fuel constants: hydrogen 120 MJ/kg (DOE AFDC, 33.3 kWh/kg) and methane 50 MJ/kg (DOE/NREL, A Comparison of Hydrogen and Propane Fuels, April 2009, p. 2). They are chemical energy, not electrical or wheel energy. Fuel use must match the selected gas. Equilibrium range uses the capacity temperature; post-fill usable reserve uses ambient temperature after cooldown. Methane is a CNG proxy; no methane/ethane/nitrogen mixing rule is applied.

Where it applies

For the exposed non-cryogenic input ranges and single-phase gas states within the CoolProp EOS limits (hydrogen up to 1000 K; methane up to 625 K). All pressures are absolute. SAE J2601, May 2020 edition, is cited for hydrogen fueling context and the nominal-density SOC convention only. The chosen cutoff ratio, 85 °C flag and predicted duration are not a qualified fueling schedule; check against the standard in force. J2601 is not a CNG protocol.

Bulk gas temperature can be below a local liner or jet hot spot. A uniform effective wall cannot reproduce radial thermal penetration or material-interface temperatures. Calibrate CwC_w, hih_i and hoh_o to the actual tank, especially for small drones, large truck tanks and long fills; fixed coefficients also limit the comparison at twice the duration. No strength, fatigue, burst, relief, permeation or dispenser-control checks are made.

At the default 320 maximum steps, doubling the resolution is checked for temperature and pressure convergence. Curves and animation interpolate the solved time history with shape-preserving cubic interpolation; no spatial temperature gradient is claimed. The section has illustrative 6 mm liner and 35 mm wrap thicknesses, independent of wall capacity and structural design. Geometry sets gas volume and thermal area; the section is fitted to the card. Pressure-class rows are comparisons of the selected gas at the capacity temperature, not vessel certifications. NIST property checks share the reference EOS family; the independently tabulated adiabatic case tests the caloric balance. The nonzero-heat-transfer comparisons use published CFD endpoints, not measured hydrogen temperatures. Zhao's model uses conjugate 2D wall conduction, spatial gas flow and a Redlich–Kwong EOS; this tool uses a uniform wall, a capsule approximation based on the published aspect ratio, equal heat-transfer areas and the Leachman EOS. Those comparisons do not establish tank-specific experimental accuracy or resolve the paper's local hot spots.

Before any fill runs, a conservative work screen estimates the thermal substeps for all three scenarios from the gas mass change, heat capacities and inner/outer conductances. Combinations above 6,000 estimated work units are rejected immediately with guidance to shorten the fill or use an offline stiff solver. This limits the supported combinations, without coarsening the accepted model or changing its physics. A separate 20,000-substep backstop detects incomplete integrations. Animated dashes indicate mass entering the valve; their spacing is illustrative and is not a prediction of local jet velocity.

References

  • CoolProp 8.0.0; Bell et al. (2014), DOI: 10.1021/ie4033999. https://coolprop.org/fluid_properties/fluids/Hydrogen.html
  • Leachman, Jacobsen, Penoncello & Lemmon (2009), Fundamental Equations of State for Parahydrogen, Normal Hydrogen, and Orthohydrogen. DOI: 10.1063/1.3160306.
  • Setzmann & Wagner (1991), methane reference EOS. DOI: 10.1063/1.555898. https://coolprop.org/fluid_properties/fluids/Methane.html
  • Yang (2009), A thermodynamic analysis of refueling of a hydrogen tank. DOI: 10.1016/j.ijhydene.2009.06.015. https://www.nist.gov/publications/thermodynamic-analysis-refilling-hydrogen-tank
  • Woodfield, Monde & Takano (2008), Heat Transfer Characteristics for Practical Hydrogen Pressure Vessels Being Filled at High Pressure, JTST 3(2), 241–253. DOI: 10.1299/jtst.3.241. https://www.jstage.jst.go.jp/article/jtst/3/2/3_2_241/_article
  • SAE J2601_202005, Fueling Protocols for Light Duty Gaseous Hydrogen Surface Vehicles, revised May 2020 (method citation only). DOI: 10.4271/J2601_202005. https://saemobilus.sae.org/standards/j2601_202005-fueling-protocols-light-duty-gaseous-hydrogen-surface-vehicles
  • U.S. DOE, Physical Hydrogen Storage. https://www.energy.gov/cmei/fuels/physical-hydrogen-storage
  • NIST Chemistry WebBook SRD 69, hydrogen and methane thermophysical tables, retrieved 8 November 2026. https://webbook.nist.gov/chemistry/fluid/ ; exact query URLs are included in the attached adiabatic reference.
  • DOE AFDC, Fuel Properties Comparison: hydrogen LHV and CNG composition. https://afdc.energy.gov/fuels/properties
  • DOE/NREL, A Comparison of Hydrogen and Propane Fuels, DOE/GO-102009-2813, April 2009, p. 2: methane LHV 50.0 MJ/kg. https://docs.nlr.gov/docs/fy09osti/45408.pdf
  • Zhao, Wei, Peng, Feng & Jia (2022), Experimental and Numerical Research on Temperature Evolution during the Fast-Filling Process of a Type III Hydrogen Tank, Energies 15, 3811, §4.3.3, pp. 12–13. DOI: 10.3390/en15103811. https://www.mdpi.com/1996-1073/15/10/3811

Inputs

InputSymbolUnitDefault
Stored gasgasHydrogen (normal H₂)
Internal volumeV_tankm^30.125
Nominal full pressure
Use 35/70 MPa (350/700 bar) for hydrogen or 20/25 MPa for CNG; every pressure is absolute.
p_fullMPa70
Capacity temperatureT_storedegC15
Minimum usable pressurep_minMPa2
Fuel used per 100 kmconsumptionkg1
Initial tank pressurep_startMPa2
Initial gas and wall temperatureT_startdegC25
Ambient temperatureT_ambientdegC25
Pre-cooled gas temperature
Temperature at the location selected under Fill boundary · advanced.
T_feeddegC-40
Station supply pressurep_stationMPa90
Pressure cutoff / nominal
A chosen model stop; 1.25 is a hydrogen study case, not a CNG filling allowance.
cut_ratio1.25
Planned fill duration
Constant inflow would deliver the nominal 15 °C mass in this time; pressure cutoff can stop it earlier.
fill_times180
Fill playback timets150
Tank length / diameter
A cylinder with hemispherical ends sets the heat transfer area.
aspect3.2
Effective wall heat capacity
Thermally active liner and composite mass × specific heat; this needs tank-specific calibration.
C_wallkJ/K80
Gas-to-wall coefficienth_insideW/(m^2*K)500
Wall-to-ambient coefficienth_outsideW/(m^2*K)4.5
Integration steps
Fourth-order Runge–Kutta maximum step; extra substeps control thermal stiffness.
N_steps320
Fill mass basisdelivery_modeNominal mass at 15 °C
Mass to addm_chargekg3
Pre-cooling temperature location
A fixed station temperature gives constant inlet enthalpy; a fixed tank-inlet temperature gives pressure-dependent enthalpy.
temperature_siteStation, before throttling

Results

With the default inputs above, the calc gives:

ResultValue
Real-gas capacity5.022 kg
Ideal-law capacity7.362 kg
Compressibility factor1.466
Ideal-law mass bias46.6 %
Equilibrium usable mass4.814 kg
Usable LHV energy160.5 kWh
Equilibrium fuel range481 km
Mass after fill5.022 kg
Peak bulk gas temperature67.7 °C
Final bulk gas temperature67.7 °C
Final wall temperature54.1 °C
Pressure after fill83.095 MPa
State of charge at cutoff100 %
Actual fill time180 s
Pressure after ambient cooldown72.502 MPa
Fuel range after fill482 km
Fill endpointPlanned mass delivered

Checked against sources

Each case below was calculated by this calc and agrees with the value from its source (within the tolerance where one is shown).

NIST hydrogen at 350 bar, 15 °C

Source: NIST Chemistry WebBook SRD 69, isotherm 288.15 K, 35 MPa: density 23.99478 kg/m³, internal energy 2521.899 kJ/kg. Independent published property service, same underlying Leachman EOS family; tests implementation and units, not a separate EOS experiment.

Inputs: Nominal full pressure = 35 MPa; the others as above.

ResultSourceThis calc
Real-gas full density23.995 kg/m³23.995 kg/m³
Full specific internal energy2,521,899 J/kg2,521,903 J/kg

NIST hydrogen at 700 bar, 15 °C

Source: NIST Chemistry WebBook SRD 69, isotherm 288.15 K, 70 MPa. Independent tabulation of density and caloric state; same reference EOS family.

Inputs: the default inputs above.

ResultSourceThis calc
Real-gas full density40.172 kg/m³40.172 kg/m³
Full specific internal energy2,488,955 J/kg2,488,959 J/kg

NIST methane at 250 bar, 15 °C

Source: NIST Chemistry WebBook SRD 69 methane isotherm 288.15 K, 25 MPa; independent tabulation, same reference EOS family.

Inputs: Stored gas = CNG proxy (pure methane); Nominal full pressure = 25 MPa; Station supply pressure = 35 MPa; Fuel used per 100 km = 4 kg; Pre-cooled gas temperature = 15 °C; the others as above.

ResultSourceThis calc
Real-gas full density199.3 kg/m³199.3 kg/m³
Full specific internal energy544,931 J/kg544,930 J/kg

Adiabatic real-gas fill against NIST caloric tables

Source: Independent endpoint obtained from NIST WebBook tabulated initial density/internal energy, supply enthalpy, and a 40.1722 kg/m³ final isochore. Algebraic first law (Yang 2009, Eq. 10), linear interpolation across a 0.5 K NIST table; no calculator engine used. It tests adiabatic real-gas heating and units, not a tank-specific heat-transfer calibration.

Inputs: Gas-to-wall coefficient = 0 W/(m²·K); Wall-to-ambient coefficient = 0 W/(m²·K); Station supply pressure = 110 MPa; Pressure cutoff / nominal = 1.5; the others as above.

ResultSourceThis calc
Final bulk gas temperature124.5 °C124.5 °C
Pressure after fill96.934 MPa96.934 MPa

Zhao 2022: 8 g/s fill with gas–wall heat transfer

Source: Zhao et al. (2022), Energies 15, 3811, DOI 10.3390/en15103811: Table 1 (145 L, 77 kg, 381 mm × 1800 mm), Table 3 (wall cp 902–920 J/kg·K), Table 4 Case 3 and §4.3.3 (180 s, 5 MPa, 273 K ambient/initial, 253 K tank-inlet temperature). Reported final temperatures and pressures are on p. 12, Figure 10 on p. 13. Whole-wall capacity uses the sourced lower bound 77×902 = 69.454 kJ/K; h_i=500 W/m²·K comes independently from Woodfield 2008 (not fitted), h_o=6 comes from Zhao §2.3. Geometry is a capsule approximation using the published aspect ratio. Paper uses conjugate 2D CFD and Redlich–Kwong; this model uses a lumped wall and Leachman EOS. Comparison tolerances 3 K and 0.6 MPa allow that difference; this is a CFD cross-check, not experimental calibration. Supply pressure is a nonbinding ceiling only because temperature is specified after throttling.

Inputs: Internal volume = 0.145 m³; Nominal full pressure = 35 MPa; Initial tank pressure = 5 MPa; Initial gas and wall temperature = -0.15 °C; Ambient temperature = -0.15 °C; Pre-cooled gas temperature = -20.15 °C; Station supply pressure = 60 MPa; Pressure cutoff / nominal = 1.5; Tank length / diameter = 4.724; Effective wall heat capacity = 69.454 kJ/K; Wall-to-ambient coefficient = 6 W/(m²·K); Fill mass basis = Specified mass increment; Mass to add = 1.44 kg; Pre-cooling temperature location = Tank inlet, after throttling; the others as above.

ResultSourceThis calc
Final bulk gas temperature21.25 ± 3 °C19.221 °C
Pressure after fill19.36 ± 0.6 MPa19.227 MPa
Accumulated gas mass1.44 kg1.44 kg

Zhao 2022: 16 g/s fill with gas–wall heat transfer

Source: Zhao et al. (2022), Energies 15, 3811, DOI 10.3390/en15103811: Table 1 (145 L, 77 kg, 381 mm × 1800 mm), Table 3 (wall cp 902–920 J/kg·K), Table 4 Case 3 and §4.3.3 (180 s, 5 MPa, 273 K ambient/initial, 253 K tank-inlet temperature). Reported final temperatures and pressures are on p. 12, Figure 10 on p. 13. Whole-wall capacity uses the sourced lower bound 77×902 = 69.454 kJ/K; h_i=500 W/m²·K comes independently from Woodfield 2008 (not fitted), h_o=6 comes from Zhao §2.3. Geometry is a capsule approximation using the published aspect ratio. Paper uses conjugate 2D CFD and Redlich–Kwong; this model uses a lumped wall and Leachman EOS. Comparison tolerances 3 K and 0.6 MPa allow that difference; this is a CFD cross-check, not experimental calibration. Supply pressure is a nonbinding ceiling only because temperature is specified after throttling.

Inputs: Internal volume = 0.145 m³; Nominal full pressure = 35 MPa; Initial tank pressure = 5 MPa; Initial gas and wall temperature = -0.15 °C; Ambient temperature = -0.15 °C; Pre-cooled gas temperature = -20.15 °C; Station supply pressure = 60 MPa; Pressure cutoff / nominal = 1.5; Tank length / diameter = 4.724; Effective wall heat capacity = 69.454 kJ/K; Wall-to-ambient coefficient = 6 W/(m²·K); Fill mass basis = Specified mass increment; Mass to add = 2.88 kg; Pre-cooling temperature location = Tank inlet, after throttling; the others as above.

ResultSourceThis calc
Final bulk gas temperature33.45 ± 3 °C33.941 °C
Pressure after fill37.66 ± 0.6 MPa37.73 MPa
Accumulated gas mass2.88 kg2.88 kg

Increment fill starting above nominal mass

Source: Regression example supplied by the independent reviewer in the user report, 8 November 2026: hydrogen, nominal pressure 35 MPa, initial pressure 40 MPa, station pressure 60 MPa, specified increment 0.1 kg, other inputs at defaults. Reviewer endpoint: 41.71 MPa, below the 43.75 MPa cutoff; pressure tolerance 0.05 bar allows rounding to 0.01 MPa. The prescribed mass addition and 180 s duration check completion without pressure cutoff. This is a behavior regression case supplied by the reviewer, not independent thermodynamic or experimental validation.

Inputs: Nominal full pressure = 35 MPa; Initial tank pressure = 40 MPa; Station supply pressure = 60 MPa; Fill mass basis = Specified mass increment; Mass to add = 0.1 kg; the others as above.

ResultSourceThis calc
Accumulated gas mass0.1 kg0.1 kg
Actual fill time180 s180 s
Pressure after fill41.71 MPa41.713 MPa

Related Resources

Responses

Your email address will not be published. Required fields are marked *

L ading...