Skip to content
Tutorials Center
My learning

Topics

Lesson 46 of 51 Ā· Projects and Practice

Project: A Reactor Kinetics Simulator

Build a reactor point-kinetics model with temperature feedback and decay heat in Python, step by step, and explore why reactors with negative feedback are self-limiting.

In this project you will build a small but realistic model of how reactor power responds to a change in reactivity. You will start from the point-kinetics equations from the lesson on reactors, add a heat balance and a temperature feedback, and then simulate a control-rod step, a shutdown with decay heat, and a comparison of feedback strengths. It is a teaching model, not a plant safety code, but it shows the same behavior that real reactors are designed around.

A block diagram: control rods give an external reactivity, which is added to feedback reactivity, drives point kinetics, gives power, heats the core, and the core temperature feeds back through a negative coefficient
The model is a feedback loop: reactivity changes power, power changes temperature, temperature changes reactivity.

What you will learn

  • How to turn the point-kinetics equations into a runnable simulation with a stiff ODE solver.
  • How to add a lumped heat balance and temperature feedback.
  • How decay heat is modeled after a shutdown.
  • How the strength of the feedback coefficient shapes the response.
  • How to check a model against simple analytic limits.

Key terms

Point kinetics: reactor physics model that treats the whole core as one point with a single neutron population n(t).

Precursors C(t): fission products whose later decay emits delayed neutrons.

Temperature coefficient α_T: change of reactivity per kelvin; negative in well-designed reactors.

Scram: rapid shutdown by inserting all control rods.

Way-Wigner formula: a standard approximation for decay-heat power after shutdown.

The model

We track three variables: the relative power n (1 means full power), the precursor concentration C, and the core temperature T.

Text
dn/dt = ((ρ āˆ’ β)/Ī›) n + Ī» C
dC/dt = (β/Ī›) n āˆ’ Ī» C
C_th dT/dt = Pā‚€ n āˆ’ h (T āˆ’ T_cool)
ρ = ρ_ext + α_T (T āˆ’ Tā‚€)

with β = 0.0065, generation time Ī› = 10⁻⁓ s, one effective precursor group with Ī» = 0.08 s⁻¹. The heat balance has full power Pā‚€, a heat-removal coefficient h and a heat capacity C_th. We choose h so that full power raises the core 500 K above the coolant, and a thermal time constant C_th/h = 10 s. The feedback coefficient is α_T = āˆ’2.5Ɨ10⁻⁵ K⁻¹, that is, āˆ’2.5 pcm per kelvin.

Analytic check. After a small step ρ_ext, the feedback must eventually cancel it: α_T Ī”T = āˆ’Ļ_ext, so Ī”T = ρ_ext/|α_T| and the new relative power is n = 1 + Ī”T/500. For ρ_ext = 0.003 that gives Ī”T = 120 K and n = 1.24. Your simulation should converge to these.

Step 1: a reactivity step with feedback

Python Step 1: control-rod withdrawal with and without feedback

Press Run (or Ctrl + Enter) to see the result here.

Program input Ā· what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Step 2: scram and decay heat

After all control rods are inserted (a large negative reactivity, say āˆ’5%), the fission power drops to about a tenth almost at once (the prompt drop, by a factor near β/(β āˆ’ ρ)) and then falls over tens of seconds as the delayed neutrons die away. But decay heat from fission products remains. The Way-Wigner formula gives it as a fraction of the pre-shutdown power, for an operating time T_op:

Text
P_decay / Pā‚€ = 0.066 [ t⁻⁰·² āˆ’ (t + T_op)⁻⁰·² ]        (t in seconds)
Python Step 2: shutdown and decay heat

Press Run (or Ctrl + Enter) to see the result here.

Program input Ā· what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Step 3: how strong must the feedback be?

Repeat the step test for several feedback coefficients, and also for a reactivity above β (prompt critical) to see how much the feedback strength limits the peak. For the prompt-critical case we cut the run short, since the model is a simplification (it ignores fuel melting, the delay of the Doppler effect and many other things).

Python Step 3: comparing feedback strengths

Press Run (or Ctrl + Enter) to see the result here.

Program input Ā· what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Questions to explore

  • How does the response change if the thermal time constant is 2 s instead of 10 s (a fast-responding fuel)?
  • What happens to the final power if the heat-removal capacity h drops by 10% during the step?
  • Try a positive feedback coefficient (α_T > 0): can you see the thermal runaway? Why would a real design avoid it?
  • Add a second delayed-neutron group (for example β₂ = 0.0013, λ₂ = 3 s⁻¹) and compare.

Common misconceptions

"If the model has feedback, a reactor cannot go wrong." Feedback limits power only if the coolant, heat removal and design stay within their assumed conditions. A model with one temperature is a simplification.

"A step reactivity above β is always catastrophic." Above β the reactor is prompt critical and the response is very fast, but how large the excursion becomes depends on feedback, geometry and shutdown systems. That is exactly why designs never allow it.

"A simulation proves a design is safe." Safety analysis uses validated, detailed codes and experiments. This project is for understanding.

Check your understanding

Quick check Ā· 5 questions

of 5 answered You scored / 5 Ā· perfect!

  1. 1. In the model, why does the power settle at a new steady level after a reactivity step?

    α_T Ī”T = āˆ’Ļ_ext at the new equilibrium, so the net reactivity returns to zero.
  2. 2. Which value of α_T gives the smallest final temperature change for the same reactivity step?

    Ī”T = ρ_ext/|α_T|, so the strongest feedback needs the smallest temperature rise.
  3. 3. One hour after shutdown from long operation, the decay heat is about...

    The Way-Wigner formula gives roughly 1% at one hour.
  4. 4. Why does the model use a stiff solver (Radau)?

    Time scales of 10⁻⁓ s and 10 s together make the equations stiff.
  5. 5. A reactivity above β means the reactor is...

    The chain reaction no longer needs delayed neutrons.

Summary

  • Point kinetics plus a heat balance and a temperature coefficient gives a compact reactor model.
  • With negative feedback the power settles where α_T Ī”T = āˆ’Ļ_ext.
  • After a scram, fission power falls quickly but decay heat (about 6% at once, about 1% after an hour) needs cooling.
  • Stronger negative feedback means a smaller power and temperature change for the same disturbance.
  • Models are for understanding; real safety work uses validated codes and experiments.

Next: Project: Hunting for the Higgs Bump