Moderator-to-Fuel Ratio Calculator
Compute the resonance escape probability, thermal utilization, and optimal moderator-to-fuel ratio for thermal reactor lattice design.
⚛️ What is the Moderator-to-Fuel Ratio?
The moderator-to-fuel ratio (M/F ratio or R) is the number of moderator atoms (or molecules) per fuel atom in a thermal reactor lattice. It is one of the most consequential design parameters in reactor physics because it simultaneously controls two competing quantities: the resonance escape probability p and the thermal utilization factor f. Both p and f depend strongly on R, and their product p×f passes through a maximum at an optimal ratio called R_opt. Maximizing p×f maximizes the component of k∞ that is directly controlled by lattice geometry, making M/F ratio optimization a central activity in nuclear fuel assembly design.
In a light water reactor (LWR), the moderator is ordinary water and R typically ranges from 1 to 5 H2O molecules per uranium atom, depending on the fuel pin pitch and pellet diameter. In a CANDU heavy-water reactor, the D2O moderator has such low neutron absorption that R can reach 300-500 D2O molecules per uranium atom, which is why CANDU runs on unenriched natural uranium. In a graphite-moderated reactor (such as the historic Chicago Pile-1 or UK Magnox stations), R must be even larger because graphite has a low slowing-down power, requiring a very large moderator volume per fuel rod to achieve adequate resonance escape.
The ratio can be expressed in several equivalent ways depending on context. Atom ratio (N_M/N_F) is used in homogeneous reactor theory and in this calculator. Volume ratio (V_M/V_F) is more common in lattice physics codes and refers to the volumes of moderator and fuel in the unit cell. For a LWR lattice, converting between atom ratio and volume ratio requires the molar masses and densities of H2O and UO2; the two ratios are related by R_atom = R_vol × (rho_M/M_M) / (rho_U/M_U). A typical PWR with a volume ratio of 1.5 corresponds to an atom ratio of about 2-3 H2O per U.
Understanding M/F ratio is essential for interpreting the moderator temperature coefficient (MTC) of reactivity. LWRs are deliberately designed slightly under-moderated (R below R_opt), so that an increase in temperature reduces water density, lowers R further below R_opt, and decreases k. This negative MTC is a passive safety mechanism: the reactor self-limits its power increase without any active control action. An over-moderated reactor would have a positive MTC, meaning temperature increases would boost reactivity and could lead to a runaway excursion.
📐 Formula
Resonance Escape Probability: p(R) = exp(−A / R)
A = I_eff / (ξ_M × σ_s,M) — parameter combining the effective resonance integral and moderator slowing-down cross-section
I_eff = effective resonance integral for U-238 (barns): 277 b for infinite dilution; typically 10-30 b for a heterogeneous LWR lattice
ξ_M = mean logarithmic energy decrement per neutron-moderator collision (H₂O: 0.920; D₂O: 0.509; graphite: 0.158)
σ_s,M = microscopic scattering cross-section of moderator per atom/molecule (H₂O: 49.2 b; D₂O: 10.6 b; graphite: 4.74 b)
R = N_M / N_F, the atom ratio of moderator to fuel
Thermal Utilization: f(R) = σ_a,F / (σ_a,F + R × σ_a,M)
σ_a,F = thermal absorption cross-section of fuel per U atom = e × 678 + (1−e) × 2.73 barns, where e is the U-235 mass fraction
σ_a,M = thermal absorption cross-section of moderator (H₂O: 0.664 b; D₂O: 0.001 b; graphite: 0.0035 b)
Optimal Ratio: R_opt = [A + √(A² + 4Aσ_a,F/σ_a,M)] / 2
Derived by setting d(p × f)/dR = 0 and solving the resulting quadratic equation in R.
k∞ estimate: k∞ = η × ε × p(R) × f(R)
Example (H₂O, 3.5% enriched, I_eff=25 b, R=2): A=0.553, p=0.759, f=0.952, p×f=0.722, R_opt=4.97
📖 How to Use This Calculator
Steps
1
Choose a calculation mode - use Evaluate Ratio to see how p, f, and k∞ behave at a specific moderator-to-fuel ratio, or Find Optimum to get the exact R_opt that maximizes p×f analytically.
2
Select moderator and fuel parameters - pick H₂O, D₂O, or graphite, set the enrichment percentage, and adjust the effective resonance integral. The default I_eff of 25 barns is appropriate for a heterogeneous LWR lattice; use 277 barns for the homogeneous infinite-dilution limit.
3
Set eta and epsilon for k∞ - enter the reproduction factor and fast fission factor. Default values of eta=2.07 and epsilon=1.05 suit a 3.5% enriched UO2 LWR assembly. Adjust these if you are studying different fuel compositions.
4
Enter the M/F ratio (Evaluate mode) - drag the slider or type a value for N_M/N_F. Typical ranges: H₂O reactor 0.5-10, D₂O reactor 50-500, graphite reactor 100-3000. The deviation readout immediately shows whether the design is under- or over-moderated.
5
Read and interpret results - the calculator shows p, f, p×f, k∞ estimate, R_opt, and the deviation from optimum. A negative deviation means under-moderated (the safer side for LWR). Click Find Optimum to instantly see the best possible p×f for the selected fuel and moderator combination.
💡 Example Calculations
Example 1 - Typical PWR Operating Point (H₂O, 3.5% Enriched, R=2)
Light water reactor with 3.5% enriched UO2, I_eff=25 b, eta=2.07, epsilon=1.05, R=2.0
1
A = I_eff / (ξ × σ_s) = 25 / (0.920 × 49.2) = 25 / 45.26 = 0.553
2
p = exp(−0.553 / 2.0) = exp(−0.276) = 0.7587
3
σ_a,F = 0.035 × 678 + 0.965 × 2.73 = 23.73 + 2.63 = 26.36 barns
4
f = 26.36 / (26.36 + 2.0 × 0.664) = 26.36 / 27.69 = 0.9520
5
p×f = 0.7587 × 0.9520 = 0.7223; k∞ = 2.07 × 1.05 × 0.7223 = 1.570
6
R_opt = (0.553 + √(0.306 + 4×0.553×26.36/0.664)) / 2 = (0.553 + 9.39) / 2 = 4.97; deviation = (2.0/4.97 − 1) × 100 = −59.8%
p = 0.7587, f = 0.9520, p×f = 0.7223, k∞ = 1.570, R_opt = 4.97 (under-moderated by 60%)
Try this example →Example 2 - Find the Optimal M/F Ratio (H₂O, 3.5% Enriched)
Find R_opt that maximizes p×f for 3.5% enriched UO2 in light water
1
A = 0.553 (same as Example 1)
2
discriminant = A² + 4A × σ_a,F / σ_a,M = 0.306 + 4 × 0.553 × 26.36 / 0.664 = 0.306 + 88.1 = 88.4
3
R_opt = (0.553 + √88.4) / 2 = (0.553 + 9.402) / 2 = 4.977
4
p_opt = exp(−0.553 / 4.977) = exp(−0.111) = 0.8948; f_opt = 26.36 / (26.36 + 4.977 × 0.664) = 26.36 / 29.67 = 0.8885
5
(p×f)_opt = 0.8948 × 0.8885 = 0.7951; k∞_opt = 2.07 × 1.05 × 0.7951 = 1.729
R_opt = 4.977, p = 0.8948, f = 0.8885, (p×f)_max = 0.7951, k∞_opt = 1.729
Try this example →Example 3 - Heavy Water CANDU-Style (D₂O, Natural Uranium, R=400)
D₂O moderated natural uranium (0.72% enrichment) at a high moderator ratio R=400
1
A = 25 / (0.509 × 10.6) = 25 / 5.395 = 4.633
2
σ_a,F = 0.0072 × 678 + 0.9928 × 2.73 = 4.882 + 2.711 = 7.593 barns
3
p = exp(−4.633 / 400) = exp(−0.01158) = 0.9885
4
f = 7.593 / (7.593 + 400 × 0.001) = 7.593 / 7.993 = 0.9500
5
p×f = 0.9885 × 0.9500 = 0.9390; k∞ = 1.34 × 1.02 × 0.9390 = 1.283 (using eta=1.34, epsilon=1.02 for natural U)
p = 0.9885, f = 0.9500, p×f = 0.9390, k∞ = 1.283 (critical-capable with natural U)
Try this example →❓ Frequently Asked Questions
What is the moderator-to-fuel ratio and why does it matter for reactor design?
The moderator-to-fuel ratio R = N_M / N_F is the number of moderator atoms per fuel atom in the reactor lattice. It controls both the resonance escape probability p (which increases with R) and the thermal utilization factor f (which decreases with R). Their product p×f passes through a maximum at an optimal ratio R_opt, making M/F ratio one of the most important tunable parameters in fuel assembly design. Getting R right determines whether the reactor can sustain a chain reaction and how it responds to temperature changes.
Why are light water reactors designed to be slightly under-moderated?
LWRs operate at R below R_opt so that the moderator temperature coefficient of reactivity is negative. When reactor power increases, the coolant heats up and its density drops, reducing R further and lowering k. This self-limiting behavior is a passive safety feature that does not require control rods or operator action. An over-moderated reactor would have a positive temperature coefficient (more power raises R toward R_opt, increasing k further), which is considered unsafe for a commercial power reactor.
What is the formula for the optimal moderator-to-fuel ratio?
Setting the derivative d(p×f)/dR equal to zero and solving gives R_opt = [A + sqrt(A² + 4A × sigma_a_F / sigma_a_M)] / 2. The parameter A = I_eff / (xi_M × sigma_s_M) combines the effective resonance integral with the moderator's slowing-down properties. For H2O and 3.5% enriched fuel with I_eff=25 barns, R_opt is approximately 5 H2O molecules per U atom. For D2O and natural uranium, R_opt is around 200-400 D2O molecules per U atom.
How does the effective resonance integral affect the calculation?
The effective resonance integral I_eff determines the parameter A = I_eff / (xi × sigma_s). A larger A means a steeper drop in p at low R, shifting R_opt to a higher value and reducing the achievable p for a given R. The infinite-dilution value for U-238 is 277 barns, appropriate for homogeneous mixtures. In heterogeneous fuel pins, self-shielding reduces I_eff to 10-30 barns. Using the wrong I_eff can significantly misplace R_opt, so it is important to use a value appropriate for the fuel geometry.
Why can CANDU reactors run on natural uranium while PWRs cannot?
The key difference is the moderator. Heavy water absorbs roughly 660 times fewer thermal neutrons per molecule than light water (sigma_a = 0.001 barns vs 0.664 barns). This means f stays high even at the very large M/F ratios required to achieve adequate p with natural uranium (0.72% U-235). In an LWR the water absorbs so many neutrons that the lattice requires at least 3% enrichment to achieve k∞ greater than 1. The extra cost of D2O production for CANDU is offset by the elimination of costly uranium enrichment.
What nuclear constants are used in this calculator for each moderator?
H2O: mean log energy decrement xi=0.920 per molecule, scattering cross-section sigma_s=49.2 barns per molecule, absorption sigma_a=0.664 barns per molecule. D2O: xi=0.509, sigma_s=10.6 b, sigma_a=0.001 b. Graphite: xi=0.158 per C atom, sigma_s=4.74 b, sigma_a=0.0035 b. For fuel: thermal absorption sigma_a_F = e × 678 + (1-e) × 2.73 barns per U atom, where e is the U-235 atom fraction and 678 barns is the 2200 m/s absorption cross-section of U-235 and 2.73 barns is that of U-238.
How does fuel enrichment change the optimal ratio?
Higher enrichment increases sigma_a_F because U-235 (678 barns) absorbs far more strongly than U-238 (2.73 barns). A larger sigma_a_F shifts R_opt upward: the fuel can dominate absorption against a larger moderator inventory before f drops too far. For natural uranium (0.72% U-235) in H2O with I_eff=25 b, R_opt is roughly 1-2, which is too low for adequate p. Enriching to 3.5% raises sigma_a_F from about 6 to 26 barns and pushes R_opt up to about 5, allowing a physically realizable LWR lattice.
What is the slowing-down power and how does it affect reactor design?
Slowing-down power = xi × Sigma_s (cm-1 units in bulk material). A high value means neutrons decelerate rapidly through the resonance energy region, giving less time for U-238 capture and a higher p. H2O has the highest SDP of common moderators, which is why LWRs are compact. Graphite has a low SDP, so graphite reactors must be physically large to achieve the same p. D2O has an intermediate SDP but compensates with very low absorption, giving an excellent moderating ratio (SDP / Sigma_a) of about 5670, compared to H2O at 68 and graphite at 170.
What is the difference between atom ratio and volume ratio for M/F?
The atom ratio R_atom = N_M / N_F counts moderator atoms per fuel atom. The volume ratio R_vol = V_M / V_F is the volume of moderator divided by the volume of fuel in the unit cell. They are related by R_atom = R_vol × (rho_M / M_M) / (rho_F / M_F), where rho and M are density (g/cm³) and molar mass (g/mol). For H2O (rho=1, M=18) and UO2 (rho=10.4, M=270), R_atom = R_vol × (1/18) / (10.4/270) = R_vol × 1.443. A volume ratio of 1.5 corresponds to an atom ratio of about 2.16.
How is k-infinity related to k-effective for a real reactor?
k∞ is the multiplication factor for an infinitely large reactor with no neutron leakage. k_eff = k∞ × P_NL, where P_NL is the non-leakage probability. For a large power reactor, P_NL is typically 0.95-0.99. For a critical reactor, k_eff = 1.000, so k∞ must exceed unity by a margin of 1-5%. Fresh fuel starts with k∞ of 1.3-1.7; control rods and boron are used to bring k_eff to 1.000. As fuel depletes over the fuel cycle, k∞ falls until k_eff can no longer reach 1, and the fuel is discharged.
Can this model be applied to fast reactors or only thermal reactors?
This model applies strictly to thermal (moderated) reactors. The formulas p = exp(-A/R) and f = sigma_a_F / (sigma_a_F + R × sigma_a_M) assume that neutrons slow down to thermal energies and are absorbed there. Fast reactors deliberately avoid moderation: there is no distinct thermal group, no resonance escape probability in the same sense, and no thermal utilization factor. Fast reactor analysis uses multigroup diffusion or transport methods with cross-sections averaged over the fast neutron spectrum, which differs fundamentally from the single-group homogeneous model used here.
How does this calculator handle moderator absorption in the denominator of f?
The thermal utilization formula f = sigma_a_F / (sigma_a_F + R × sigma_a_M) accounts for all thermal neutron absorbers in the denominator. In this simplified model only fuel and moderator are included; structural materials, control poisons, and fission products are not. Adding these extra absorbers (each with their own macroscopic absorption cross-section) would lower f further. For a first-pass lattice design this two-component model gives a useful estimate, but final design calculations require full neutron transport codes like CASMO or OpenMC with complete material inventories.