ESP Drives, Step-Up Transformers and Long Cables: Separating Resonance from Voltage Drop

1. Define the electrical boundary

An electric submersible pump may place a surface VFD and step-up transformer far from a downhole motor. The transformer ratio sets an approximate fundamental voltage relationship. It does not guarantee a safe motor-terminal waveform when converter switching, cable capacitance, leakage inductance and motor impedance interact. Define the location of the output filter, transformer and measurement points before requesting a resonance assessment.

SLB's ESP power-system portfolio includes variable-speed equipment and application-specific cables. The engineering implication is to request coordinated interface data from the actual suppliers, not combine individual catalog ratings into an assumed system approval. ESP system scope.

2. Obtain frequency-dependent data

Request transformer leakage impedance with its frequency and test basis, winding capacitances where relevant, losses, permitted converter waveform, minimum operating frequency and volts-per-hertz limits. Record ratio and impedance at the intended tap. A line-frequency nameplate impedance is not a complete high-frequency transformer model.

For the cable, request length, capacitance and resistance per unit length, temperature assumptions, conductor arrangement, armor/ground return and splice details. Include motor winding surge limits, phase-to-ground insulation, motor lead extension, drive carrier range and filter capacitance, damping and supported settings. Downhole temperature changes resistance and thermal margin. Treat differential-mode and common-mode paths separately; their return circuits are different.

3. An illustrative resonance screening

For a deliberately simplified single equivalent mode referred to one side of the transformer, assume 2 mH leakage inductance and 2 km cable with 0.20 µF/km effective capacitance. This is a hypothetical lumped model, not an ESP product dataset. The capacitance is defined for this equivalent mode; do not mix phase-to-ground and phase-to-phase values.

Suppose the proposed drive carrier range is 4–6 kHz. The illustrative resonance lies in that range, which triggers a detailed review; it does not prove failure at that frequency. Cable distributed effects, motor loading, winding capacitance, filter placement and damping can shift or suppress the response. A lower switching frequency is not automatically a solution because sidebands and other modes remain possible.

4. Match the remedy to the mechanism

Danfoss distinguishes differential sine filtering from common-mode and combined filtering and publishes product compatibility. Its filter overview also explains why smoothing phase-to-phase voltage does not automatically solve cable leakage. These are useful distinctions, not a blanket approval for a kilometer-scale ESP installation. Filter modes; output-filter overview.

5. Keep low-frequency duty in the same review

Check fundamental voltage drop at hot cable resistance, motor voltage across speed, transformer excitation and filter reactive current. Raising a tap to compensate cable drop may overvolt equipment at lighter load or another speed. Keep RMS heating, peak voltage, dv/dt and common-mode voltage as separate acceptance quantities. A satisfactory RMS reading can coexist with harmful high-frequency stress.

Do not treat an output sine filter as an input harmonic filter. Their purposes and connection points differ. The harmonic impedance-scan guide discusses the related principle of frequency-dependent networks; ESP studies require their own output-side model. Input supply disturbances belong in the ride-through guide.

6. Procure a coordinated evidence package

Assign one party responsibility for interface approval. Require the modeled cable range, drive and firmware, permitted carrier/fundamental combinations, transformer tap and waveform limits, filter tolerances and motor insulation envelope. Ask for voltage and current predictions at named terminals under minimum and maximum cable length, speed and loading, plus ground-fault behavior.

Commission through an approved OEM procedure using suitable bandwidth and isolation for the quantities being measured. Surface terminal data do not directly measure downhole motor peaks; explain the validated model linking them. Retain settings and cable configuration in the well record, and reassess extensions or replacement drives rather than carrying forward the old approval.

7. Questions and references

Does a sine filter remove every resonance? No. Mode, placement, damping and supported operating range matter.

Can transformer impedance at mains frequency predict switching peaks? Not by itself; parasitic and distributed parameters matter.

Does a correct RMS motor voltage prove insulation adequacy? No. Peak and common-mode stresses must also be evaluated.

Primary references: SLB ESP systems; Danfoss filter modes; Danfoss output-filter overview.