Theory: Feedback, Thermal Limits, and Decay Heat
Why Feedback Matters
A power reactor is not held critical by its control rods alone. As power rises, the coolant and fuel heat up, and that heating changes k. Designers make sure these reactivity coefficients are negative, so that any rise in power or temperature removes reactivity and the reactor tends to return to where it was. This inherent stability is required by NRC General Design Criterion 11. It also means that operators must add positive reactivity, usually by diluting soluble boron, to raise power and keep the reactor critical.
Moderator and Fuel Temperature Coefficients
The moderator temperature coefficient (MTC) describes how reactivity changes with coolant temperature. Hot water is less dense, so the lattice becomes less moderated; in an under-moderated PWR this lowers p more than it raises f, and the MTC is negative. Dissolved boron leaves the core with the expanding water, which pushes the MTC towards positive, so the MTC is least negative at the beginning of cycle (high boron) and most negative at end of cycle (low boron). The fuel temperature (Doppler) coefficient comes from the broadening of U-238 resonances as fuel heats, which increases resonance capture and lowers p. It acts instantly, because the fuel heats before the coolant does.
Power defect = Δρ_mod + Δρ_Doppler (from hot zero power to power P)
Boron change to stay critical: Δppm = power defect / boron worth
Core Thermal Power from the Neutron Flux
Every fission releases about 200 MeV, of which about 3.2 × 10⁻¹¹ J is recovered as heat in the core. The fission rate per cubic centimetre is Σ_fφ, so the thermal power of the whole core follows from the core-average flux and volume.
q′_avg = P / (N_rods × L_rod) (W/cm)
Power Peaking Factors
In a bare cylinder the flux falls from the centre to the edges as a Bessel function J₀ radially and a cosine axially. The peaking factor is the ratio of the peak to the average. The hot spot sets the thermal limits, so the core must be designed and operated so that the peak linear heat rate stays below the technical specification limit (typically about 18 kW/ft, 590 W/cm) and fuel centreline melting and cladding damage are avoided. Fuel managers flatten the power with low-leakage loading patterns (fresh fuel interspersed with burned fuel, burned fuel on the periphery) and burnable absorbers; in-core detectors measure the real peaking factors.
F_Q = F_R × F_Z q′_max = F_Q × q′_avg
Decay Heat and Shutdown Cooling
When a reactor trips, fission stops within seconds, but the fission products already in the fuel keep decaying. Their beta and gamma energy, the decay heat, is about 6 to 7% of full power just after shutdown, about 1% after an hour, and still a few tenths of a percent after a month. It must be removed continuously; losing it was the cause of fuel damage at Three Mile Island and Fukushima Daiichi. The Way–Wigner correlation gives a good estimate. The residual heat removal (RHR) system carries the heat to a heat exchanger.
Q = ṁ c_p ΔT → ṁ = P_d / (c_p ΔT)
Data Used in This Lab
| Quantity | BOC | EOC |
|---|---|---|
| Soluble boron at full power | about 1200 ppm | about 50 ppm |
| Moderator temperature coefficient | −10 pcm/°C | −45 pcm/°C |
| Doppler (fuel temperature) coefficient | −3 pcm/°C | −3 pcm/°C |
| Differential boron worth | −8 pcm/ppm | −10 pcm/ppm |
| Core data | Value | Core data | Value |
|---|---|---|---|
| Rated thermal power | 3411 MWt | Core radius; height | 170 cm; 366 cm |
| Fuel rods; active length | 50 952; 365.8 cm | Core-average Σ_f | 0.10 cm⁻¹ |
| Energy recovered per fission | 3.2 × 10⁻¹¹ J | Linear heat rate limit | 590.6 W/cm (18.0 kW/ft) |
| T_avg programme | 291.7 + 0.161 P °C | Specific heat of coolant | 4.2 kJ/kg·°C |
Values are typical of a four-loop Westinghouse-type PWR and follow the DOE Fundamentals Handbook (Nuclear Physics and Reactor Theory; Heat Transfer, Fluid Flow and Thermodynamics) and the NRC Reactor Concepts Manual. Real coefficients vary with burnup, boron, and power, and are supplied for each cycle in the plant's core operating limits report.
Operations
Reactor operators dilute boron as power rises and borate as it falls, and track the power defect on every power change.
Reactor engineering
Engineers check the peaking factors from in-core flux maps against the core operating limits report every month.
Safety analysis
Decay-heat curves size the RHR, emergency core cooling, and spent fuel pool cooling systems.
Planning with Decay-Heat and Power-Defect Graphs
Operators and engineers plan with graphs of these same equations. A decay-heat curve for the actual operating history shows when the decay heat falls below the capacity of one cooling train, which sets how long redundant trains or other heat sinks must stay available after shutdown. Because long-lived fission products build up only over a long run, the operating history makes little difference in the first minutes but a large one after days. A power-defect graph is used for load following: the feedback reactivity changes linearly with power, so the boron change for a planned maneuver is the power defect per percent power times the change in power, divided by the boron worth.
Apparatus
The plant systems and instruments used to manage feedback, thermal limits, and decay heat. In the simulation these are modelled for you, and the readings correspond to what each would show.
Instructions: Running the Virtual Experiment
You are the reactor engineer on shift. The simulation has three immersive 3D work areas and two graph work areas. The parts build from undergraduate work in Parts A, B, and C to graduate-level analysis in the last step of Parts D and E; the level of each part is shown beside its title. In each one you set the plant conditions, read the instruments, work out the answer yourself, enter it, and press Check. Record every reading and calculation in your lab report tables (modelled in the Example Report), with a labelled screenshot of each setup. Drag in any 3D view to orbit, and scroll to zoom. Use the data tables in the Theory section.
Simulation: Power Ascension Feedback, Core Peaking, Shutdown Cooling, and Planning Graphs
The reactor is critical at hot zero power (0%, T_avg = 291.7 °C). Raise the power. The average coolant temperature follows the programme T_avg = 291.7 + 0.161 P °C, and the average fuel temperature rises to T_fuel = T_avg + 6.0 P °C (P in %). Part A uses this work area.
Setup
Readings
Coefficients are per °C. Boron worth is negative: adding boron removes reactivity, diluting it adds reactivity.
Work it out, then check
| Cycle | Power (%) | ΔT_mod (°C) | ΔT_fuel (°C) | Δρ_mod (pcm) | Δρ_Doppler (pcm) | Power defect (pcm) | Dilution (ppm) |
|---|
A 3411 MWt-class core: radius 170 cm, height 366 cm, 50 952 fuel rods each 365.8 cm long. Set the average thermal flux and the loading pattern, work out the core power and the hot-spot linear heat rate, and compare it with the limit. Part B uses this work area.
Setup
Readings
Each fission releases 3.2 × 10⁻¹¹ J. A uniform bare core has F_R = 2.316 (Bessel J₀ shape) and F_Z = 1.571 (cosine shape).
Work it out, then check
| Loading | φ (× 10¹³) | P (MW) | q′_avg (W/cm) | F_Q | q′_max (W/cm) | Within 590 W/cm? |
|---|
The reactor has tripped from full power after a long run. Fission has stopped, but the decay of fission products keeps producing heat. Choose the time since shutdown and size the residual heat removal (RHR) flow. Part C uses this work area.
Setup
Readings
Way–Wigner decay heat correlation, valid from about 10 s to 100 days after shutdown.
Work it out, then check
| T (operation) | t (after shutdown) | P_d/P₀ | P_d (MW) | ΔT (°C) | ṁ (kg/s) |
|---|
Decay heat depends on how long the reactor ran before shutdown. After a long run, more long-lived fission products have built up, so decay heat stays higher for longer. Part D uses this work area.
Setup
Readings
Way–Wigner: P_d = 3411 × 0.0622 [ t^(−0.2) − (t + T)^(−0.2) ] MW. Read crossing times from the graph, then refine by trial.
Work it out, then check
| History | Train capacity (MW) | Time to reach capacity (h) | Time after shutdown | Ratio 3 y ÷ 30 d |
|---|
Plants that follow grid demand change power during the day. Every power change shifts the coolant and fuel temperatures, so the feedback reactivity changes and boron must be adjusted to keep the reactor critical. Part E uses this work area.
Setup
Readings
Power defect at P = (MTC × 0.161 + FTC × 6.161) × P. Reactivity change = defect(final) − defect(start).
Work it out, then check
| Cycle | Start (%) | Final (%) | Δρ feedback (pcm) | Boron change (ppm, + dilute) |
|---|
This accessible version provides the same set-up, work-out, and check exercises as the 3D and graph work areas, using native controls, data tables, and screen-reader announcements.
Team Questions
Example Lab Report
Sample report showing the expected format and level of detail. Use it as a guide for your own submission, and include labelled screenshots of each setup.
Reactivity Feedback, Power Peaking, and Decay Heat
Nuclear Engineering Technology | Section: [Your Section] | Date: [Date]
Lab Members: [Names of all members present]
Purpose
To calculate the moderator and Doppler feedback and the power defect for a power ascension at beginning and end of cycle, and the boron dilution needed to stay critical; to calculate core thermal power, average and peak linear heat rates, and compare them with the technical specification limit; to calculate decay heat after shutdown and the RHR flow needed to remove it; and to use decay-heat and power-defect graphs to plan cooling-train capacity and boron changes for load following.
Theory
Feedback is Δρ = MTC ΔT_mod + FTC ΔT_fuel, and the boron change to stay critical is the power defect divided by the boron worth. Core power is P = 3.2×10⁻¹¹ Σ_fφV, the hot-spot linear heat rate is q′_max = F_R F_Z q′_avg, and decay heat follows P_d/P₀ = 0.0622[t^(−0.2) − (t + T)^(−0.2)], removed at ṁ = P_d/(c_pΔT).
Table A: Power Defect and Boron Dilution (from 0% power) Undergraduate
| Cycle | Power (%) | ΔT_mod (°C) | ΔT_fuel (°C) | Δρ_mod (pcm) | Δρ_Doppler (pcm) | Power defect (pcm) | Dilution (ppm) |
|---|---|---|---|---|---|---|---|
| BOC | 50 | 8.05 | 308.1 | −80.5 | −924.2 | −1004.7 | 125.6 |
| BOC | 100 | 16.10 | 616.1 | −161.0 | −1848.3 | −2009.3 | 251.2 |
| EOC | 50 | 8.05 | 308.1 | −362.3 | −924.2 | −1286.4 | 128.6 |
| EOC | 100 | 16.10 | 616.1 | −724.5 | −1848.3 | −2572.8 | 257.3 |
Sample calculation (BOC, 100%): Δρ_mod = −10 × 16.10 = −161.0 pcm; Δρ_Doppler = −3 × 616.1 = −1848.3 pcm; defect = −2009.3 pcm; dilution = −2009.3 / −8 = 251.2 ppm.
Table B: Core Power and Peaking Undergraduate
| Loading | φ (× 10¹³) | P (MW) | q′_avg (W/cm) | F_Q | q′_max (W/cm) | Within 590.6 W/cm? |
|---|---|---|---|---|---|---|
| Uniform (bare) | 3.20 | 3403 | 182.6 | 3.638 | 664 | no |
| Flattened | 3.20 | 3403 | 182.6 | 2.175 | 397 | yes |
| Uniform (bare) | 2.80 | 2977 | 159.7 | 3.638 | 581 | yes |
Sample calculation: V = π × 170² × 366 = 3.323×10⁷ cm³; P = 3.2×10⁻¹¹ × 0.10 × 3.2×10¹³ × 3.323×10⁷ = 3.403×10⁹ W; q′_avg = 3.403×10⁹ / (50 952 × 365.8) = 182.6 W/cm.
Table C: Decay Heat after 1 Year at 3411 MWt (ΔT = 25 °C) Undergraduate
| t after shutdown | P_d/P₀ | P_d (MW) | ṁ (kg/s) |
|---|---|---|---|
| 10 s | 0.03728 | 127.15 | 1211.0 |
| 1 minute | 0.02546 | 86.84 | 827.0 |
| 1 hour | 0.01013 | 34.54 | 328.9 |
| 8 hours | 0.00601 | 20.50 | 195.3 |
| 1 day | 0.00444 | 15.14 | 144.2 |
| 1 week | 0.00238 | 8.12 | 77.3 |
| 30 days | 0.00131 | 4.46 | 42.5 |
Table D: Decay-Heat Planning (one cooling train) Upper undergraduateGraduate step
| History | Train capacity (MW) | Time to reach capacity (h) | Time after shutdown | Ratio 3 y ÷ 30 d |
|---|---|---|---|---|
| 1 year | 20 | 8.78 | 1 hour | 1.188 |
| 1 year | 30 | 1.79 | 1 day | 1.516 |
| 1 year | 10 | 91.99 | 30 days | 3.984 |
| 30 days | 20 | 4.14 | 1 week | 2.247 |
| 3 years | 20 | 11.32 | 8 hours | 1.350 |
Crossing times found by trial: P_d = 3411 × 0.0622 [t^(−0.2) − (t + T)^(−0.2)] = capacity. Include a screenshot of the decay-heat curves with the capacity line.
Graduate step: integrating Way–Wigner, E = 3411 × 0.0622 × [t^0.8 − (t + T)^0.8]/0.8 between t and t + 24 h. After a 1-year run the next 24 h release 1672.3 GJ starting at 1 hour and 1169.1 GJ starting at 1 day; after a 3-year run, 1786.6 GJ starting at 1 hour. 1672 GJ would boil off roughly 740 t of water at 2.26 MJ/kg.
Table E: Load-Following Power Changes Upper undergraduateGraduate step
| Cycle | Start (%) | Final (%) | Δρ feedback (pcm) | Boron change (ppm, + dilute) |
|---|---|---|---|---|
| BOC | 100 | 50 | 1004.6 | −125.6 |
| BOC | 50 | 100 | −1004.6 | 125.6 |
| BOC | 30 | 90 | −1205.6 | 150.7 |
| EOC | 100 | 50 | 1286.4 | −128.6 |
| EOC | 50 | 100 | −1286.4 | 128.6 |
| EOC | 30 | 90 | −1543.7 | 154.4 |
Power defect per % power: BOC −20.093 pcm/%, EOC −25.728 pcm/%. Δppm = Δρ / boron worth.
Graduate step: with a √T Doppler law calibrated at 100% (T_f(P) = 564.85 + 6.161P K), the feedback change is +921.1 pcm for 100% → 50% at BOC (linear: +1004.6), −1156.3 pcm for 30% → 90% at BOC (linear: −1205.6), and −847.6 pcm for 0% → 30% at EOC (linear: −771.8). The boron change for 0% → 30% at EOC becomes 84.8 ppm of dilution instead of 77.2 ppm.
Discussion
Both coefficients were negative, so raising power removed reactivity and boron had to be diluted by about 250 ppm to reach full power. The Doppler effect supplied most of the power defect, because the fuel heats by hundreds of degrees while the coolant heats by only 16 °C. At end of cycle the moderator coefficient was four and a half times more negative, because the low boron concentration no longer offsets the loss of moderation, so the power defect grew from −2009 to −2573 pcm. A large negative MTC also means a cooldown, such as a steam-line break, adds a lot of positive reactivity at end of cycle.
A uniform fresh core at rated flux put the hot spot at 664 W/cm, above the 590.6 W/cm limit, even though the average was only 183 W/cm. The flattened low-leakage loading lowered F_Q from 3.64 to 2.18 at the same power, bringing the hot spot to 397 W/cm. With a uniform loading, power would have to be held to about 2977 MW (87%). Decay heat fell from 3.7% of full power at 10 s to 1.0% at 1 hour and 0.13% at 30 days. Even after a month, 4.5 MW must still be removed, which is why shutdown cooling and spent fuel cooling run continuously.
The decay-heat curves showed that operating history matters little in the first minutes but a great deal later. At 1 hour the 3-year core made only 1.19 times the decay heat of the 30-day core, but at 30 days it made 3.98 times as much, because long-lived fission products only build up over a long run. After a 1-year run, one 20 MW train could carry the core alone from about 8.8 hours, but a 10 MW train only from about 92 hours (3.8 days). The power-defect graph showed a straight line for each part of the cycle, so the boron change for a load change is simply the power defect per percent times the change in power. Dropping from 100% to 50% needed 126 ppm of boration at BOC and 129 ppm at EOC. The EOC value is only slightly larger because the steeper power defect is largely offset by the higher boron worth.
The graduate steps refined both simple pictures. Integrating the decay-heat curve showed that the first day after a trip from a long run still releases about 1.7 TJ, enough to boil roughly 740 tonnes of water if all cooling were lost, which is why cooling is protected by redundant trains. Replacing the linear Doppler model with the √T law left the full-power defect unchanged but moved more of it to low power: the 0% → 30% change gained about 76 pcm of negative feedback, and the 100% → 50% change lost about 83 pcm. Operators therefore expect proportionally more dilution during the first part of a power ascension than a straight-line estimate predicts.
Conclusion
Negative feedback coefficients make the reactor self-limiting but require boron dilution to raise power. Power peaking, not average power, sets the thermal limit, and core loading patterns are designed to flatten it. Decay heat remains significant for weeks, and the RHR flow must be sized for it.
Practice Questions
Show all work and include units. Use the data tables in the Theory section.