Analyse the water-level oscillation of a surge tank (surge chamber) on a hydropower headrace tunnel. Adjust the tunnel length and area, the tank area and the steady velocity to see the period and maximum surge amplitude of the undamped mass oscillation after a load rejection, updating in real time.
Parameters
Headrace tunnel length L
m
Tunnel length from the reservoir to the surge tank
Tunnel cross-sectional area A_t
m²
Surge tank cross-sectional area A_s
m²
Horizontal area of the shaft. A larger tank gives a smaller surge
Steady velocity v₀
m/s
Tunnel flow velocity before the rejection
Results
—
Oscillation period T (s)
—
Max surge amplitude Z_max (m)
—
Angular frequency ω (rad/s)
—
Tunnel/tank area ratio
—
Max rate of level rise (m/s)
—
Surge magnitude
—
Hydropower layout in section — water-level animation
Reservoir, headrace tunnel, surge tank and turbine. The tank level slowly rises and falls about the reservoir level as a decaying sinusoid. The dotted line is the maximum surge.
Oscillation period T and maximum surge amplitude Z_max of the undamped mass oscillation. L is the headrace tunnel length, A_t and A_s are the tunnel and tank areas, v₀ is the steady velocity and g is gravity. A larger tank area A_s reduces the surge amplitude.
Angular frequency ω and the maximum rate of rise of the tank level. The level z(t) follows a simple harmonic z(t)=Z_max·sin(ωt), rising fastest right after the rejection.
What is a Surge Tank?
🙋
A "surge tank" is that vertical tower you always see rising from the tunnel in hydropower-plant diagrams, right? What is it actually for?
🎓
Exactly, that tower. In a hydropower plant the water travels from the reservoir to the turbine along a long, low-gradient headrace tunnel. That tunnel holds hundreds or thousands of tonnes of water flowing continuously. The problem comes when the turbine suddenly stops. When the guide vanes slam shut — a "load rejection" — that heavy water column cannot stop instantly because of its inertia. With nowhere to go, the trapped pressure runs back along the tunnel and, in the worst case, bursts it. That is "water hammer".
🙋
Wait, the tunnel bursts?! That sounds terrifying. So the surge tank prevents that?
🎓
Yes — the surge tank is the elegant cure for water hammer. You build a tall open shaft right next to the turbine, as close upstream as practical. When the flow to the turbine is suddenly cut, the rejected water runs up the surge tank instead, and its level rises. Rather than slamming a pressure surge back into the tunnel, the system converts it into an up-and-down motion of the water level. Try raising the "surge tank area" on the left — you will see the maximum surge amplitude Z_max get smaller.
🙋
And after the level rises, what happens? Does it just overflow?
🎓
No — you design the tank tall enough that it does not. Once the level has risen far enough, the weight of the built-up water (the head) pushes the tunnel flow back the other way. Then the level falls, overshoots below the reservoir level, and rises again. The whole tunnel-and-tank system oscillates like a giant U-tube of water swinging slowly back and forth. The "tank water-level oscillation" chart below is exactly that sine curve.
🙋
The period on the chart is about 200 seconds. Is the oscillation really that slow?
🎓
Yes, and that is the fascinating part of a surge tank. Water hammer itself travels at the speed of sound — a "sub-second" phenomenon — but the surge-tank mass oscillation is on the order of minutes. With T = 2π·√(L·A_s/(g·A_t)), the longer the tunnel and the wider the tank, the longer the period. Two quantities dominate the design: the period T and the maximum surge Z_max. Z_max sets how tall the tank must be built. In a real plant, friction adds damping, so the swing gets a little smaller with every cycle.
🙋
I see. So widening the tank reduces the surge, but it costs more money.
🎓
That is exactly the trade-off. Double the tank area A_s and Z_max drops by a factor of √2, but the cost of a huge vertical shaft soars. So in practice you look for "the smallest tank that keeps the surge within an acceptable range". The "maximum surge amplitude vs tank area" chart below is that falling curve. This tool uses the simplest, frictionless idealised model, but it gives you a solid first feel for surge-tank design.
Frequently Asked Questions
For an undamped, frictionless mass oscillation the period is T = 2π·√(L·A_s/(g·A_t)), where L is the headrace tunnel length, A_t is the tunnel cross-sectional area, A_s is the surge-tank cross-sectional area and g is gravity. The period grows with the tunnel length and the tank area, and shrinks with the tunnel area. For real surge tanks the period is typically minutes, not seconds, and this tool shows T in seconds.
The maximum surge amplitude Z_max is the peak height the tank water rises above the steady level after a sudden full load rejection, given by Z_max = v0·√(L·A_t/(g·A_s)), where v0 is the steady velocity. It dictates how tall the surge tank must be built. A larger tank cross-section A_s reduces the amplitude, but at greater construction cost. In a real tunnel friction damps the oscillation, so each successive swing is smaller than the last.
A hydropower plant carries water from the reservoir to the turbines along a long, low-gradient headrace tunnel. That tunnel holds an enormous mass of moving water, and when the turbine guide vanes close suddenly — a load rejection — the water column cannot stop instantly because of its inertia. With nowhere to go, a violent pressure surge (water hammer) would slam back along the tunnel and could burst it. The surge tank is a tall open shaft placed close to the turbine; it harmlessly diverts the rejected water upward and converts the surge into a slow rise and fall of the water level.
When the load rejection cuts the flow to the turbine, the water with nowhere to go runs up the surge tank and the level rises. Once it has risen high enough, the built-up head reverses the tunnel flow; the level then falls, overshoots below the reservoir level, and rises again. The whole tunnel-and-tank system behaves exactly like a giant U-tube of water oscillating slowly back and forth. This tool models that mass oscillation in its simplest, undamped form.
Real-World Applications
Hydropower headrace tunnel design: This is the most basic use of a surge tank. In pondage and reservoir hydropower plants with a long pressure tunnel, a surge tank is placed just upstream of the turbine to absorb the water hammer of a load rejection. The designer first uses the undamped mass oscillation to get a rough period and maximum surge, then adds friction losses, the turbine response time and simultaneous rejection of several units before settling the tank height and area.
Pumped-storage plants: A pumped-storage plant has long waterways on both the upper and lower reservoir sides, with frequent switching between generation and pumping and frequent rapid stops. Surges arise not only from a load rejection in generating mode but also from a sudden stop in pumping mode, so surge tanks or surge chambers are often provided on both the upstream and downstream sides. The mass-oscillation period and amplitude also influence the operating schedule.
Long-distance water conveyance and pump stations: Beyond hydropower, long-distance water-supply and irrigation pipelines also use surge tanks (standpipes) or surge towers to suppress the pressure transients caused by a sudden pump stop. The longer the pipeline and the higher the velocity, the larger the surge, so the tank size can be estimated with the same approach this tool uses.
Preliminary study of differential and orifice surge tanks: In real plants, the surge is damped faster by throttling a port (orifice) or by adopting a differential (riser-type) surge tank. Before that detailed design, an undamped model like this tool is used first to pin down "what order the oscillation period is" and "how tall the tank needs to be", and to judge whether it is worth adding damping and non-linear effects.
Common Misconceptions and Pitfalls
The biggest pitfall is confusing the surge-tank mass oscillation with water hammer. They are phenomena on completely different time scales. Water hammer is a "sub-second" high-speed phenomenon in which a pressure wave travels at the speed of sound (about 1400 m/s in water) — an elastic pressure transient. The surge-tank mass oscillation handled by this tool, on the other hand, is a slow "minutes" phenomenon in which the water in the tunnel and tank moves back and forth almost as a rigid body. The very role of the surge tank is to "convert" fast water hammer into a slow mass oscillation and render it harmless. If you do not study both separately, you will get the design wrong.
Next, assuming the frictionless, undamped model is the real answer for an actual plant. The Z_max = v0·√(L·A_t/(g·A_s)) used here is an idealised value with zero friction loss. A real headrace tunnel has wall friction, which damps the mass oscillation. Friction has two effects: one is that the oscillation shrinks with each cycle and eventually settles at the steady level; the other is that it makes the first upsurge itself smaller than the undamped value (although the down-surge is not always smaller, because the friction direction reverses). In real design, recomputing with a model that includes friction is essential, and the undamped value should be regarded only as a conservative estimate.
Finally, the misconception that "you only need to look at the first upsurge for the maximum surge". The first upsurge after a load rejection is often the largest, but the design must also consider turbine restarts and cases where several surges superimpose. For example, if the turbine is re-engaged the moment a surge has fallen below the reservoir level, the next surge can add to the previous one and produce a level swing larger than the first. The surge-tank height must be decided against "the most severe surge among all plausible operating sequences", and you must not judge the margin from a single isolated load rejection alone.
How to Use
Enter tunnel length (lNum) in metres—typical range 500–5000 m for hydropower headrace tunnels
Set surge tank diameter (atNum) in metres; larger cross-sectional area reduces oscillation amplitude
Input tunnel cross-sectional area (asNum) in m²; ratio atNum/asNum directly affects damping characteristics
Specify initial water velocity (vNum) in m/s representing steady-state flow before load rejection
Run simulation to generate oscillation period T (s), maximum surge amplitude Z_max (m), and angular frequency ω (rad/s)
Worked Example
Consider a hydropower scheme with headrace tunnel length 1800 m, tunnel area 3.5 m², surge tank diameter 2.8 m (area 6.16 m²), and initial steady flow velocity 0.95 m/s. The simulator calculates oscillation period T ≈ 8.4 seconds, maximum surge amplitude Z_max ≈ 1.7 m above equilibrium, and angular frequency ω ≈ 0.748 rad/s. The area ratio (6.16/3.5 = 1.76) indicates moderate tank-to-tunnel coupling; rate of level rise peaks at approximately 0.38 m/s during initial transient.
Practical Notes
Surge tank diameter critically influences stability: doubling atNum reduces Z_max by roughly 40% in typical schemes
Longer tunnels (lNum > 3000 m) produce slower oscillations; check T remains compatible with turbine governor response time (usually 2–6 seconds)
Tunnel friction damping neglected here; real amplitude decays 10–25% per cycle depending on roughness and Reynolds number
Maximum level rise rate governs tank freeboard; ensure Z_max + safety margin (0.5 m minimum) does not exceed intake depth