Electronic Speed Control of Three Pumps on a Drilling Rig

When rotation, the mud pump and the compressor are all driven by the same engine, a single throttle lever cannot manage three jobs at once. An application note on the flow-speed relationship, the annular velocity threshold, NPSH margin, cable hysteresis and closed loop feedback.

On a drilling rig, a single throttle lever cannot manage three jobs at once. Rotation, mud circulation and feed are generally driven from the same diesel engine through hydraulic circuits; engine speed is a single variable, but it determines all three functions. This application note explains the physical reasoning behind three independent speed channels and what to watch for during installation.

The problem: one variable, three different optima

Balancing all three with one lever means holding the engine near full throttle for the needs of the most demanding group - the other two groups then run at unnecessarily high speed. On truck-mounted drilling rigs, an arrangement in which the rig and the chassis share the same diesel engine and the pumps are PTO- or belt-driven is common in the industry; engine monitoring, emergency stop and electronic throttle control are standard components on the control panels.

The relationship between flow and speed

In a positive displacement pump, Q = D x n x ηv; D is displacement (volume swept per revolution/stroke), n is speed and ηv is volumetric efficiency. In a mud pump, displacement is found from π x (liner diameter / 2)² x stroke length x number of cylinders.

With D and ηv held constant, Q ∝ n: flow is a linear function of speed, and any percentage error in speed appears in flow at the same percentage. To see this, a representative calculation (it does not belong to a real machine; it is there to show how the formula is used; D = 0.80 L/rev, ηv = 0.90):

Going from 1200 to 1600 rpm increases flow by 33.3% - exactly the same proportion as the increase in speed. Reading it in the other direction is more instructive: if you are not going to deviate more than 5% from the target flow, speed must be held to better than 5% accuracy. The friction and backlash band of a mechanical throttle lever makes it hard to get below that.

Mud circulation: why is flow so critical?

Drilling fluid controls formation pressures, removes cuttings from the hole, seals permeable formations, cools and lubricates the bit, transmits hydraulic energy to downhole tools and - perhaps most importantly - maintains borehole wall stability.

The quantity that determines cuttings transport performance is annular velocity: the upward velocity of the mud through the annular cross section between the drill string and the borehole wall.

An important note on preference: it is preferable to achieve hole cleaning by raising annular velocity rather than viscosity, because high viscosity reduces the penetration rate. The maximum achievable annular velocity is in turn limited by pump capacity, bottomhole pressure and borehole wall stability.

When the operator says "the cuttings aren't coming up", the first variable he reaches for is mud pump speed. Being able to do that steplessly and repeatably is the most concrete benefit of independent potentiometer control.

Why does steady speed matter?

Pulsating pressure and fatigue

Triplex mud pumps produce pulsating flow; the successive acceleration-deceleration cycle of the fluid is the primary cause of pressure peaks. With every pulse those peaks accelerate the deterioration of the pump itself, of fluid-end consumables and of the equipment downstream of the pump. A pulsation dampener that is faulty or under-charged allows the full pressure pulses to be transmitted into the surface lines. Continuous oscillation in speed (hunting) changes the amplitude and frequency content of the pulses; steady speed removes that variability.

Cavitation and NPSH

Cavitation occurs when available NPSH falls below required NPSH. The critical point is this: required NPSH increases with speed, while available NPSH falls because inlet line losses grow as flow increases. So raising the speed narrows the safety margin from both directions. On a pump with a weak suction side (a long or dirty suction hose, low pit level), "giving it a little more throttle" can push it past the cavitation threshold. Being able to lock the speed at a precise and repeatable point makes it possible to work below that threshold with a safe margin.

The hysteresis problem in cable-pull control

Control cables exhibit lost motion and friction; using only the command signal causes the lever to hunt continuously within the cable slack. Hysteresis generally comes from shaft-bushing wear or backlash in the actuator coupling, and it is direction-dependent: the butterfly settles in a slightly different place depending on which side of the actuator's friction band it pressed against last.

In mechanical governors, the deadband is the speed range in which the governor does not respond because of friction and backlash, and it is typically 0.3-1.0% of rated speed.

The result: in an open loop drive, a "50% command" corresponds to one position on the way up and a different one on the way down. Remembering that Q ∝ n, that difference directly degrades flow repeatability.

The solution: closed loop position feedback

The GS740-3P measures actuator shaft position with an AS5600 contactless magnetic encoder and compares it to the command. Whatever the tension and slack in the cable, the controller brings the actual shaft angle to the target; direction-dependent hysteresis is confined inside the control loop.

Over a restricted angle range the step count is found from N = (θ_max / 360) x 4096:

(The angle values are examples and do not belong to a specific machine.) Even a 45° stroke means 512 discrete steps - a position resolution finer than 0.2% of the working range. Compared with the mechanical governor deadband (0.3-1.0%), the sensor resolution stays below the deadband: the bottleneck is not the sensor but the mechanics - and closed loop control is the right tool to correct that.

Why not a potentiometer or a micro-switch?

One distinction matters: the three adjustment potentiometers in the operator's hand are a user interface element - a component inside the cabin, moved infrequently, with a low duty cycle. Where the service-life problem lies is the position feedback, which moves continuously and sits in the vibration, heat and dirt around the engine block; that is where contactless measurement is used.

Calibration: why a wire instead of a button?

The GS740-3P has no idle-to-working transition button; calibration is done by briefly touching the blue memory wire to black (GND), and the upper limit that is taught rescales the working range to 0-100%.

An interesting confirmation: one of the angle programming options in the AS5600's own data sheet is exactly this topology - teaching the start and stop angle by pulling a pin to ground for at least 100 ms and releasing it. The data sheet further states that in applications that do not use the full turn, programming the range used causes the output resolution to scale automatically to that range.

In a drilling cabin it is often impossible to reach a button on the engine; a physical button also means a panel opening that needs IP protection and a mechanical part that can fail. Wire contact eliminates both risks at once.

Wiring and noise: where does the shield connect?

Why is the shield connected to green and not to black? The shield connects to the point where the reference of the protected signal is. The reference of the yellow signal is the green line (potentiometer GND), not the power chassis. Connecting the shield to power GND mixes the voltage drop along the line carrying motor return current into the signal's reference. Connecting the shield to both green and black creates exactly the grounded-at-both-ends condition and causes noise current to flow through the shield.

What to do in the field:

1. The shield is connected to green only at the control unit end; at the potentiometer end it is cut and insulated.
2. The cut end of the shield is not left long - it behaves like an antenna.
3. Yellow (signal) and green (reference) run together inside the same jacket.
4. The power line is not run alongside the signal cable over long distances; where they must cross, they cross at a right angle.

Field safety: where is the limit?

Drilling carries a particular hazard: when the rig accidentally penetrates a pocket of natural gas, the escaping gas can mix into the diesel engine's intake air and the engine can run away uncontrolled even if the fuel supply is cut. US offshore regulation (30 CFR § 250.405) requires diesel engines used on drilling rigs to be equipped with an air intake device to stop the engine in a runaway.

Critical conclusion: In a runaway, closing the throttle lever will not stop the engine - the engine is taking its fuel from outside. No electronic speed control unit, this product included, can be presented as runaway protection or take the place of an air shutoff valve. The emergency stop chain must also remain independent.

The fail-safe value of leaving the mechanical lever in place

Overcurrent and stall protection are part of the same philosophy: when the actuator runs into a mechanical obstruction (a seized butterfly shaft, a broken cable conduit), the current limit engages and protects both the actuator and the throttle mechanism.

System summary

For single-pump installations the GS740 is used: the working speed is stored by holding the IP67 metal button for 3 seconds, and a short press switches between idle and working speed.

Related content

• Guide to Electronic Speed and Throttle Control on Diesel Engines

Knowledge Center