SVC vs. STATCOM: A US Engineering Guide to Selecting Dynamic Reactive Power Compensation

November 12, 2026 | Blog

Part A — SVC vs. STATCOM at a Glance

When a US transmission planner, a renewable developer or a large-load customer needs fast, continuously controllable reactive power, the shortlist almost always starts with two shunt FACTS devices: the static var compensator (SVC) and the static synchronous compensator (STATCOM). A third option, the synchronous condenser, belongs on the same list whenever system strength or inertia is part of the problem.

All three regulate voltage by injecting or absorbing reactive power at a bus. They differ in how they produce that reactive power. That difference decides how much support each delivers when the grid needs it most, during and just after a fault, when voltage is depressed. It also decides how fast each responds, how much harmonic filtering it needs, how much land it occupies, what it costs to own for 30 years, and how it behaves on a weak grid crowded with inverter-based resources (IBRs).

This guide is Keentel Engineering's head-to-head selection guide. It does not repeat the voltage-stability and motor-stall mechanics covered in our earlier STATCOM article. Instead it compares the two technologies on the questions that decide a procurement:

• Topology: what is actually inside each device.
• Physics: the exact equations behind their voltage–current behavior.
• Response time: what "fast" really means.
• Harmonics: what each injects and how it is filtered.
• Ownership: losses, footprint and lifecycle cost.
• Weak grids: how each behaves at low short-circuit strength.
• US compliance: where each fits under NERC and IEEE.
  
  

In plain terms: an SVC is a bank of reactors and capacitors whose effective size is adjusted by thyristor switches, so its output depends on the square of voltage. A STATCOM is a power-electronic voltage source that pushes a controlled current, so its output depends only on voltage itself. Both regulate voltage well in normal operation; they separate sharply when voltage collapses.

The two technologies side by side

Six findings from Keentel's analysis

• The V-versus-V² difference is real, but often overstated. A STATCOM does not "maintain full reactive power regardless of voltage." At its current limit its output falls in proportion to voltage: 50 Mvar at 0.5 pu for a 100 Mvar unit, against 25 Mvar for an SVC. The advantage is a factor of 1/V. That is decisive in deep sags and modest at 0.9 pu, where the numbers are 90 vs. 81 Mvar.
  
• Response time is mostly about the control loop, not the semiconductor. Thyristors and IGBTs both switch in microseconds to milliseconds. The closed-loop voltage response that a planner actually sees is set by measurement filtering, regulator gain, slope and system strength. A well-tuned SVC can be fast enough; a poorly tuned STATCOM can be slow or unstable.
  
• The SVC's harmonic filters are part of its design, not an add-on. A TCR's 5th harmonic peaks at about 5% of its rated current. The filters that absorb it also supply part of the capacitive range, so they must be engineered against the system's harmonic impedance from the outset
  
• Lifecycle cost differences are smaller than capex differences. In our illustrative ±100 Mvar comparison, the STATCOM costs about 40% more up front. Over 30 years at 7%, it ends up about 34% higher in present-value terms, after lower losses and O&M are credited. Land, outage cost and schedule can close the rest of that gap at constrained sites.
  
• On weak grids neither device adds system strength. Where short-circuit ratios fall toward 2 or below, the binding problem is often grid strength itself, which only synchronous condensers, grid-forming inverters or new transmission can supply.
  
• The right answer is often a hybrid. Mechanically switched capacitors (MSC) or reactors (MSR) handle slow, steady-state needs, so a smaller fast device can be reserved for dynamic events. This applies to both SVC- and STATCOM-based solutions.
  

Figure 1. Left: idealized voltage–current characteristics with a 3% slope. In the control range both devices behave identically; below it, the SVC is bounded by its maximum susceptance while the STATCOM holds rated current. Right: maximum capacitive output of 100 Mvar units as bus voltage falls. The shaded area is the extra reactive support a STATCOM delivers during a sag.

Engineering review of common claims

The illustrative system used in this guide



Keentel's worked examples use one consistent setting so the numbers can be compared directly.

Part B — SVC vs. STATCOM: An Engineering Selection Guide

1. Why dynamic reactive power matters

Voltage on a transmission system is held up by reactive power. Generators, shunt capacitors, line charging and FACTS devices supply it; loads, transformer and line reactance, and stalled motors consume it. In steady state, slow resources are enough: switched capacitor banks and reactors, transformer tap changers, and generator voltage schedules. During disturbances they are not. A fault depresses voltage across a wide area for a few cycles. After it clears, the system must re-establish voltage within a fraction of a second, or induction motors stall, inverter-based plants enter low-voltage modes and protection begins to operate.

Dynamic reactive devices fill that gap. They change output continuously and fast, holding a voltage set point through load swings and supporting recovery after faults. US planners need them for four recurring reasons:

• Post-contingency voltage support to meet TPL-001-5.1 steady-state and transient voltage criteria, especially in load pockets and long radial paths.
  
• Voltage stability margin where heavy transfers push a path near the nose of its PV curve.
  
• Generator and IBR interconnection requirements, including the 0.95 power factor requirement of FERC Order No. 827 and dynamic reactive support under IEEE 2800-2022 as adopted by transmission owners.
  
• Fluctuating loads: arc furnaces, large motors, rolling mills and, increasingly, AI data centers whose demand swings within seconds.

Mechanically switched capacitors (MSC) and reactors (MSR) are the cheapest reactive resource per Mvar, but they switch in discrete steps, take several cycles to close, need discharge time before reclosing and wear with each operation. SVCs and STATCOMs provide continuous, repeatable, fast control. The choice between them is the subject of the rest of this guide.

2. How an SVC works

2.1 Building blocks

An SVC is an assembly of shunt reactors and capacitors whose effective susceptance is controlled by thyristor valves. It connects to the transmission bus through a step-down coupling transformer, so the valves and passive components operate at a medium voltage. In the US that medium voltage is typically in the 13.8–34.5 kV class, and our examples use 34.5 kV. The main elements are:

• Thyristor-controlled reactor (TCR): an air-core reactor in series with a pair of anti-parallel thyristors in each phase. Delaying the firing angle reduces the conduction period and therefore the effective inductive susceptance, continuously from full rating to zero.
  
• Thyristor-switched capacitor (TSC): a capacitor bank in series with anti-parallel thyristors and a small damping reactor. A TSC is either fully on or fully off; it is never phase-controlled, because that would cause large inrush transients.
  
• Fixed capacitors and harmonic filters: capacitor banks with series reactors tuned near the TCR's dominant harmonics (5th, 7th, sometimes 11th and 13th). They absorb TCR harmonics and supply part of the capacitive range.
  
• Mechanically switched capacitors or reactors (MSC/MSR): breaker-switched banks on the medium-voltage or transmission bus, used for slow adjustments under the SVC controller's supervision.
  
• Controls and protection: voltage measurement, a regulator with slope, firing logic, valve protection and coordination with external banks.

2.2 Common topologies

New ParagraTCR branches are usually delta-connected. With balanced firing, triplen harmonic currents (3rd, 9th, 15th) circulate inside the delta and do not appear in the line currents. That is why the line-current spectrum is dominated by the 5th, 7th, 11th and 13th.

2.3 TCR susceptance as a function of firing angle

The firing angle convention matters, because textbooks use two. In this guide the firing angle α is measured from the zero crossing of the voltage across the TCR branch. Full conduction occurs at α = 90° (firing at the voltage peak), and the reactor is blocked at α = 180°. The conduction angle is σ = 2(π − α). With this convention, the fundamental-frequency susceptance of a TCR with inductance L is:

Equivalently, in terms of the conduction angle:

If a reference measures α from the voltage peak instead (0° = full conduction, 90° = blocked), substitute α' = α − 90°. Mixing the two conventions is a common source of specification and modeling errors, so a technical specification should state the convention explicitly.

Worked Example 1 — Sizing the TCR and TSC for a ±100 Mvar SVC at 138 kV

Requirement. The SVC must deliver +100 Mvar capacitive to −100 Mvar inductive at the 138 kV bus. For simplicity, ignore coupling-transformer reactive losses (in a real design they add several Mvar and shift the medium-voltage ratings).

Topology. A TSC-TCR design with fixed harmonic filters:

• Fixed filters (5th and 7th tuned branches): 30 Mvar capacitive, always connected.
• Two TSC steps of 35 Mvar each: 70 Mvar capacitive.
• One delta-connected TCR.

Capacitive limit. Filters plus both TSCs, with the TCR blocked: 30 + 70 = 100 Mvar capacitive.

Inductive limit. Both TSCs off, the filters still connected, the TCR at full conduction. To achieve −100 Mvar net, the TCR must absorb 100 + 30 = 130 Mvar at rated voltage.



TCR branch reactance. Each delta branch carries one third of the rating at the 34.5 kV line voltage:

That corresponds to L = 72.9 mH per branch. The full-conduction line current is 130 MVA / (√3 × 34.5 kV) = 2,176 A.

Susceptance versus firing angle. Using the equation above:

The relationship is strongly nonlinear: half the TCR's range lies between 90° and about 120°. The SVC controller therefore linearizes it, computing the firing angle that produces the susceptance its voltage regulator demands.

Filling the gaps. With 35 Mvar TSC steps and a 130 Mvar TCR, every net output between −100 and +100 Mvar can be reached continuously. The controller selects the number of TSCs in service and sets the TCR to absorb the excess. Overlap between steps (TCR range larger than one TSC step) gives hysteresis, so the TSCs do not chatter.

2.4 TSC transient-free switching

A capacitor energized at an arbitrary point on the voltage wave draws a large inrush current. A TSC avoids this by firing its thyristors only when the voltage across the valve is near zero. That happens when the system voltage equals the voltage left on the capacitor, typically at a voltage peak when the capacitor is fully charged. Once blocked, the capacitor holds its peak charge, so it can be re-inserted within about one cycle without a transient. This is why TSC steps can switch much faster and more often than mechanically switched capacitors, which need several minutes of discharge time before reclosing.

2.5 SVC voltage–current characteristic

Within its control range an SVC is regulated to a slope, also called droop or current compensation:

Here I is positive for inductive current, and the slope reactance X(SL) is typically 1–5% on the SVC's rating base. Outside the control range the SVC is simply a fixed susceptance at its limit, so its current and reactive power are

3. How a STATCOM works

3.1 The voltage-source converter principle

A STATCOM is a voltage-source converter connected to the grid through a coupling reactance: a phase reactor, the leakage reactance of a coupling transformer, or both. The converter synthesizes a three-phase voltage Vc, nearly in phase with the system voltage V. The reactive power exchanged through the coupling reactance X is:

When Vc exceeds V, the STATCOM supplies reactive power (capacitive operation). When Vc is below V, it absorbs reactive power (inductive operation). A small phase shift lets the converter draw enough real power from the grid to cover its losses and hold its DC capacitor voltage. No large energy store is needed for reactive operation, because reactive power circulates between the phases.

In practice the converter is controlled as a current source. An inner current loop in a rotating dq reference frame, synchronized to the grid by a phase-locked loop (PLL), sets the reactive current in milliseconds. An outer voltage loop with slope sets the reactive current reference.

3.2 Converter topologies

An MMC STATCOM in delta configuration with dozens of submodules per arm produces an output voltage staircase so fine that harmonic filters are often unnecessary. It is also modular: redundant submodules let it keep running after individual failures, which improves availability.

3.3 STATCOM voltage–current characteristic

In its control range a STATCOM uses the same slope characteristic as an SVC. Outside it, the converter is current-limited:

Two practical details matter in specifications:

• Short-term overload. Many STATCOMs can deliver 1.2 pu or more current for a few seconds, limited by semiconductor junction temperature. This capability must be specified, tested and modeled; it is not automatic.
• Low-voltage behavior. Below a threshold, often 0.2–0.3 pu, the converter may block or reduce current because the PLL cannot track a collapsed voltage reliably. A STATCOM's support therefore does not continue all the way to zero voltage.

3.4 DC capacitor sizing in brief



The DC capacitor must limit ripple and ride through transients, not store energy for reactive output. A useful first estimate relates stored energy to converter rating through a time constant:

Converters typically use τ of a few milliseconds to tens of milliseconds. In an MMC the energy is distributed across submodule capacitors, sized mainly for the voltage ripple each submodule sees. Adding substantial storage (a battery) turns the device into a STATCOM with energy storage, which can also exchange real power, for example for fast frequency response.

4. Reactive output when voltage collapses

Worked Example 2 — Support from 100 Mvar units at depressed voltage

At its capacitive limit, an SVC is a fixed capacitor: Q = B(max)V². A STATCOM at its current limit is a fixed current source: Q = V·I(max). For 100 Mvar units rated at 1.0 pu voltage:

Interpretation. In the voltage range where most planning criteria live, 0.9–1.05 pu, the two devices are nearly equivalent: an SVC rated about 11% higher matches a STATCOM at 0.9 pu. The STATCOM's advantage grows sharply during deep sags, which is exactly when fault-induced delayed voltage recovery (FIDVR) and motor stalling begin. A planner who needs support at 0.7 pu would need an SVC rated about 143 Mvar to match a 100 Mvar STATCOM. That is the correct way to compare them: equal performance at the voltage that matters, not equal nameplate.

Two cautions keep the comparison honest:

• The 0.2 pu row is theoretical for many STATCOMs. If the controls block below 0.3 pu, the STATCOM delivers nothing at 0.2 pu. Ask the supplier for the actual low-voltage characteristic and verify it in models.
  
• Capacitive support is not the whole story. After fault clearing, voltage can overshoot, especially with capacitors and filters connected. An SVC's TCR and a STATCOM's inductive range must both absorb reactive power quickly to limit temporary overvoltage.

5. Response time: device versus closed loop

Specifications often quote response in a single number. It helps to separate three layers:

The closed-loop response is set by the voltage regulator's gain, which must be tuned to the system's short-circuit strength. The plant seen by the regulator is approximately:

A regulator gain that gives a crisp response on a strong system becomes too aggressive when the system weakens after line outages: the effective loop gain rises as S(sc) falls. Both SVCs and STATCOMs therefore use gain supervision or adaptive gain, which reduces gain when oscillation is detected or when system strength is estimated to be low. A specification should require stable response across the full range of credible system strengths, not just a fast step at the strongest one.

For most transmission voltage-support needs, a 2–5 cycle response is fast enough. TPL-001 transient voltage criteria and FIDVR recovery play out over hundreds of milliseconds to seconds. Speed becomes decisive for flicker compensation, where the load changes within a cycle, and for some ride-through and power-oscillation-damping applications.

6. System strength and weak-grid behavior

Worked Example 3 — How much does 100 Mvar move the voltage?

Use ΔV ≈ ΔQ/S(sc) at the 138 kV bus, with the short-circuit ratio taken against a 300 MW plant at the same

On a strong system the device nudges voltage; on a weak one it dominates it. Three consequences follow:

• Gain scheduling is essential. A gain tuned for 2,000 MVA is four times too aggressive at 500 MVA.
• Rating needs shrink but sensitivity grows. Fewer Mvar are needed to correct a given voltage deviation, but control interactions with nearby IBRs become likely.
• The linear estimate fails. A 20% voltage change cannot be computed with small-signal formulas. Load-flow and EMT studies are needed.

6.1 Neither device makes the grid stronger

An SVC adds no fault current; at its limit it is just a susceptance. A STATCOM adds fault current only up to its converter rating, roughly 1.0–1.2 pu of its own current, and that current depends on a PLL tracking a severely distorted voltage. Neither provides inertia. For problems where the root cause is low system strength (IBR control instability, protection that cannot see faults, high voltage sensitivity), the candidate solutions are:

• Synchronous condensers, which provide real short-circuit current (typically 4–6 times rating for a few cycles), rotating inertia and a voltage source behind reactance.
• Grid-forming (GFM) converters, including GFM STATCOMs with energy storage, which behave as a voltage source and do not depend on a PLL in the same way.
• Transmission reinforcement, which raises short-circuit strength directly.

6.2 Comparison with synchronous condensers

The choice is not always either–or. Several US weak-grid regions are evaluating STATCOMs paired with synchronous condensers, or condensers alone, where IBR penetration has driven system strength down.

6.3 Control interactions

On weak grids, fast voltage controllers can interact with nearby IBR plant controllers, HVDC converters and each other. Common symptoms include sustained oscillations at a few hertz to tens of hertz, and voltage hunting between devices with similar time constants. Mitigations are:

• Coordinate slope and gain with nearby plant controllers, and set deliberately different response speeds.
• Run EMT studies with vendor-specific, user-defined models of every nearby converter, not only generic RMS models.
• Check the SVC's capacitive banks for resonance with the network at low system strength, where the network impedance rises.

7. Harmonics and filtering

Worked Example 4 — TCR harmonics and a tuned filter

A delta TCR with balanced firing produces characteristic harmonics of order 6k ± 1 in the line current. The magnitude of each harmonic depends on the firing angle and peaks at an intermediate angle. As a percentage of the TCR's rated fundamental current:

For the 130 Mvar TCR in Worked Example 1, with a line current of 2,176 A at full conduction, the worst-case 5th harmonic is about 0.0505 × 2,176 ≈ 110 A at 34.5 kV.



Filter design. Consider a 15 Mvar single-tuned 5th harmonic filter at 34.5 kV, tuned slightly below the harmonic at 4.7. The capacitor reactance for 15 Mvar effective output, accounting for the series reactor, is:

That gives C = 31.9 µF per phase (wye-equivalent) and L = 9.98 mH. The fundamental current is about 251 A, and the capacitor sees about 1.047 pu voltage because of the series reactor. The capacitor must be rated for that voltage plus the harmonic contribution, consistent with IEEE 18 and IEEE 1036 capacitor duties.

Why tune at 4.7, not 5.0? Tuning slightly low leaves margin for capacitor aging, temperature drift and frequency deviation. All of those shift the tuning upward. A filter tuned exactly at 5 could move above the 5th harmonic, where a parallel resonance with the system impedance sits right next to it.

Design checks. Every SVC filter design should check:

• Background harmonics already on the system, which the filter will also absorb.
• Parallel resonances between the filters and the network across the full range of system strengths and contingencies.
• Capacitor, reactor and resistor duty including harmonics.
• Compliance at the point of common coupling with IEEE 519-2022 voltage and current distortion limits.

STATCOMs. An MMC STATCOM typically produces harmonic voltages well within IEEE 519 limits without filters. A small high-pass filter may still be used, and two-level and three-level converters usually need filters for switching harmonics. The network's own background harmonics and resonance must still be studied, because the coupling transformer's capacitance and nearby capacitor banks can create resonances.

8. Losses, footprint and lifecycle cost

8.1 Losses

Losses depend strongly on topology and on how the device spends its time:

• FC-TCR SVCs have high losses near zero output, because the TCR conducts heavily to cancel the fixed capacitors.
• TSC-TCR SVCs keep losses low near zero output, because the TCR conducts lightly.
• MMC STATCOMs have losses of roughly 0.8–1.2% at rated output and much less at light load. Older designs with higher switching frequencies were around 1.5–2% or more.
• Auxiliaries (cooling, HVAC, controls) add tens to hundreds of kilowatts for any of these technologies.

The meaningful number is energy loss over the actual duty cycle. Most transmission devices sit near zero output most of the time, waiting for a disturbance, so standby losses dominate.

Worked Example 5 — 30-year lifecycle comparison

Illustrative assumptions for a ±100 Mvar device at 138 kV:

The present value factor is:

Net present cost is capex plus PVF times annual costs: for the SVC, 30 + 12.409 × 0.453 = $35.63M; for the STATCOM, 42 + 12.409 × 0.467 = $47.80M. The STATCOM is about 34% higher.

What this example leaves out:

• Land and site work. A STATCOM's footprint is often 40–60% of an equivalent SVC's. At urban or constrained sites, land can swing the decision.
• Rating at depressed voltage. If the requirement is defined at 0.7 pu, the SVC must be rated about 143 Mvar to match. That raises its cost by tens of percent and nearly closes the gap.
• Outage and schedule costs. Shorter construction and modular redundancy can be worth more than loss savings.
• Refurbishment. Thyristor valves and controls may need replacement after 20–25 years. Converters may need submodule and control replacement on a similar or shorter cycle. Spare parts and obsolescence plans belong in the specification.

All costs here are illustrative. Real quotations vary with market conditions, scope and site; use them only for structure, not budgets.

8.2 Footprint

9. Hybrid solutions

Many of the best US projects use a fast device for dynamic needs and cheaper switched banks for steady-state needs:

• STATCOM plus MSCs/MSRs: a ±50 Mvar STATCOM with 2 × 50 Mvar MSCs gives −50 to +150 Mvar of total range. The STATCOM's full dynamic range is kept available for disturbances, and the MSCs switch in slowly as steady-state conditions change.
• SVC core with MSCs: the same principle with thyristor-based control.
• Hybrid STATCOM with TSCs: a STATCOM with thyristor-switched capacitors, extending capacitive range at lower cost than converter capacity.
• STATCOM with energy storage: adds real-power capability for frequency response or oscillation damping. This blurs into battery energy storage systems (BESS) with grid-support functions.

A coordinated controller is essential. It should keep the fast device near the center of its range in steady state ("reactive reserve management") by switching the slow banks, so full dynamic range is available when a fault occurs.

10. Where each device fits in US practice

10.1 Transmission voltage support under TPL-001

Planners use dynamic reactive devices to meet TPL-001-5.1 steady-state and transient voltage criteria after contingencies, and to avoid FIDVR in load pockets. Selection hinges on the voltage at which support is needed. If post-contingency voltages fall below about 0.8 pu during recovery, the STATCOM's V-versus-V² advantage typically wins on equal performance. If the need is mainly steady-state voltage control near 0.95 pu, an SVC or a hybrid with MSCs can be the lowest-cost solution.

10.2 IBR plants: Order 827 and IEEE 2800

FERC Order No. 827 requires newly interconnecting non-synchronous generators to provide dynamic reactive power within 0.95 leading to 0.95 lagging power factor at the high side of the generator substation. IEEE 2800-2022, adopted to varying degrees by US transmission owners and regions, sets reactive capability and ride-through performance at the point of measurement. Most plants meet these with inverter capability, but shortfalls are common in long collector systems.

Worked Example 6 — Reactive shortfall at a 300 MW solar plant

A 300 MW solar plant connects at 230 kV. At 0.95 power factor, the reactive requirement at the high side is:

Losses in the collector system and the main power transformer add about 30 Mvar, so the inverters must provide about 128.6 Mvar when the plant is at full output. With 320 MVA of inverter capacity:

The shortfall is 128.6 − 111.4 = 17.2 Mvar. The two common remedies are:

• Oversize the inverters to about √(300² + 128.6²) ≈ 326.4 MVA, which also adds capability at all output levels.
• Add a ±20 Mvar STATCOM at the collector bus, which also helps with ride-through and voltage regulation.



An SVC is rarely chosen for this role, because its output falls with V² during ride-through, its filters complicate the plant's harmonic study, and its footprint is larger. Always confirm the transmission owner's specific requirements, including whether reactive capability must be available at zero active power output.

10.3 Arc furnaces and flicker

Electric arc furnaces cause rapid, erratic reactive power swings that create voltage flicker. IEEE 1453 sets flicker assessment methods and planning levels. Historically, large SVCs with fast TCRs were the standard solution. STATCOMs now offer better flicker reduction factors because of their faster response, especially at higher frequencies within the flicker range. Selection rests on the required flicker reduction, the furnace size relative to system strength, and budget.

10.4 Large loads and data centers

Large AI data centers can change demand by hundreds of megawatts within seconds, and their power electronic loads can trip offline during voltage dips. Both effects stress local voltage. In load pockets at 138 kV and 230 kV, utilities are studying dynamic reactive devices at or near large-load interconnections. A STATCOM's fast response and low-voltage performance are attractive here, but the real issue may be large-load ride-through behavior, which is a load-side engineering problem.

10.5 HVDC and series-compensated corridors

Line-commutated HVDC converters consume large amounts of reactive power and need strong AC voltage. SVCs and STATCOMs have both supported HVDC terminals. Near series-compensated lines, subsynchronous interactions must be screened for any new converter-based device.

11. Standards and compliance map

Applicability caution. A standalone STATCOM owned by a transmission owner is generally not a generating resource, and NERC's IBR registration criteria and PRC-029 apply to IBR generating facilities. When a STATCOM is part of an IBR plant, it is usually included in the plant's performance obligations. Confirm applicability with the planning coordinator and regional entity for each project.

12. Modeling for planning and design studies

Studies at each stage use different models:

• Load flow: an SVC as a voltage-controlled shunt with susceptance limits; a STATCOM as a voltage-controlled source with current limits. Many programs have dedicated SVC and FACTS device records.
• Positive-sequence stability (RMS): generic library models for SVCs and STATCOMs are available in the major US planning programs, for example the SVSMO family and the CSTCNT-type STATCOM model. Model names and features vary by program and version, so confirm the accepted models with the planning coordinator. For MOD-032 submissions, use the approved model list in the region.
• EMT: for weak grids, harmonic interactions, subsynchronous screening and control interactions, require the supplier's detailed EMT model with real control code, validated against factory tests.
• Harmonic studies: frequency-domain studies of the filters and network resonance, with the device modeled as a harmonic source.

Model validation after commissioning, by comparing recorded responses to simulations, should be part of the contract.

13. Specification checklist

A good functional specification, following the structure of IEEE 1031 and IEEE 1052, defines:

• Ratings: continuous capacitive and inductive Mvar at a defined voltage, and short-term overload capability (magnitude and duration).
• Voltage range: continuous operating range, and performance at depressed voltage down to a stated level.
• Response: closed-loop response time and overshoot for defined step tests, across a defined range of system strengths.
• Control modes: voltage control with slope, reactive power control, power oscillation damping, coordination with MSCs/MSRs and reactive reserve management.
• Harmonics: performance against IEEE 519 at the point of common coupling, including background harmonics and network impedance envelopes.
• Losses: guaranteed losses at defined operating points, with loss evaluation in the bid comparison.
• Availability and redundancy: target availability, redundant submodules or valve levels, spares.
• Environmental: ambient temperature range, altitude, seismic qualification to IEEE 693, audible noise limits.
• Models: RMS and EMT models validated against testing, with control code and documentation.
• Cybersecurity: compliance with applicable CIP requirements and utility security standards.
• Training, documentation, warranty and lifecycle support: including obsolescence management.

14. Commissioning tests

Commissioning records should be kept as the baseline for future PRC-005 maintenance, model updates and performance reviews.

15. Decision matrix

The right answer comes from studies, not rules of thumb: load flow and voltage-stability studies to size the need, dynamic studies to test performance at depressed voltage, EMT studies on weak grids, harmonic studies for any device with filters, and a lifecycle cost comparison on an equal-performance basis.

Part C — Case Studies

The following case studies are composites based on typical US project conditions. Names, locations and identifying details are withheld, and numbers are rounded.

Case Study 1 — An aging 345 kV SVC: refurbish or replace?

Situation



A transmission owner in the Midwest operated a ±200 Mvar SVC at a 345 kV substation, in service for about 28 years. The thyristor valves and the analog-era control system were near the end of their supported life. Spare parts were scarce, and the SVC had two forced outages in three years. Meanwhile, the region's generation had shifted from coal plants to wind, and post-contingency studies now showed voltages dipping to about 0.75 pu during the first second after certain 345 kV faults.

Analysis

Keentel compared three options on an equal-performance basis:

• Refurbish the SVC: new valves and digital controls, reusing reactors, capacitors and the transformer after condition assessment. Lowest cost, but the V² characteristic delivered only about 112 Mvar at 0.75 pu.
• Replace with a ±200 Mvar STATCOM: about 150 Mvar at 0.75 pu, a smaller footprint and lower harmonics, but at the highest cost.
• Hybrid: ±125 Mvar STATCOM plus 2 × 75 Mvar MSCs at 345 kV, reusing part of the existing yard. At 0.75 pu, the STATCOM delivered about 94 Mvar dynamically, with MSCs switching in within seconds for sustained needs.

Dynamic studies showed that the critical recovery period needed about 140 Mvar of fast support at depressed voltage. Refurbishment alone fell short. The hybrid met the requirement once the STATCOM's 1.2 pu short-term overload (about 112 Mvar at 0.75 pu) was combined with a fast MSC that was already in service before the contingency under reactive reserve management.

Engineered solution

The owner selected the hybrid. Keentel prepared the functional specification under IEEE 1052, including the low-voltage characteristic and overload ratings, coordinated control with the MSCs, and an EMT model requirement.

Outcome

The new installation met the transient voltage criteria with margin, freed about 40% of the yard and was cheaper than a full-size STATCOM by about a quarter. The old SVC's coupling transformer was retained as a spare for the system after testing.

Case Study 2 — A 230 kV solar plant short of reactive power

Situation

A 250 MW solar project in the Southwest, connecting at 230 kV, had a long collector system and a large main power transformer. During the facility study, the transmission owner identified a reactive capability shortfall at the high side at full output, and the IEEE 2800-based ride-through requirements also had to be met.

Analysis

Keentel's reactive capability study found a shortfall of about 14 Mvar at 0.95 lagging power factor, with collector and transformer losses accounting for most of it. Oversizing the inverters would have required renegotiating supply contracts late in the project. The site was in a moderately weak area, with an SCR near 3.

Engineered solution

A ±20 Mvar STATCOM at the 34.5 kV collector bus, coordinated with the power plant controller. The STATCOM handled fast dynamic response, while the plant controller dispatched the inverters for steady-state voltage control. EMT studies with vendor-specific inverter and STATCOM models confirmed stable operation at the minimum system strength.

Outcome

 The plant met its reactive capability and ride-through requirements without changing the inverter order. Commissioning tests matched the model responses closely, which simplified the plant's model validation submission.

Case Study 3 — Flicker at a 138 kV steel mill

Situation

A steel producer planned to add an electric arc furnace at a 138 kV service in the Southeast. Pre-project studies predicted short-term flicker severity well above the utility's planning levels under IEEE 1453 at the point of common coupling, which also served residential customers.

Analysis

Keentel modeled the furnace using measured data from a similar furnace and evaluated an SVC and a STATCOM. Both reduced flicker, but the SVC's predicted flicker reduction factor (about 2) was not enough to meet the planning level. The STATCOM's faster response achieved a factor of about 4. The SVC also required larger 2nd- and 3rd-harmonic filters to handle the unbalanced, interharmonic-rich furnace current.

Engineered solution

A STATCOM sized to the furnace's reactive swings, with a small fixed capacitor bank for power factor correction and filtering. The control was configured for fast load compensation with independent phase control.

Outcome

Measured flicker after commissioning was within planning levels, and harmonic distortion at the point of common coupling met IEEE 519. The mill's power factor improved enough to reduce demand charges, and the furnace's productivity rose because voltage at the electrodes was steadier.

Part D — 20-Question Technical FAQ

How Keentel Engineering Helps

Keentel Engineering supports utilities, independent power producers, large-load developers and industrial customers through every stage of a dynamic reactive power project:

• Needs assessment: load flow, PV/QV and dynamic studies to define how many Mvar are needed, where, how fast and at what voltage.
• Technology selection: equal-performance comparison of SVC, STATCOM, hybrid and synchronous condenser options, with lifecycle cost analysis.
• Weak-grid and EMT studies: system strength screening, control interaction studies and EMT modeling with vendor-specific models.
• Harmonic and flicker studies: filter design review, resonance scans and compliance with IEEE 519 and IEEE 1453.
• Specification and procurement: functional specifications based on IEEE 1031 and IEEE 1052, bid evaluation and loss evaluation.
• Interconnection support: reactive capability studies for IBR plants under FERC Order No. 827 and IEEE 2800, and model submissions under MOD-032.
• Substation design and protection: physical layout, grounding, protection and control design for the coupling transformer, filters and banks.
• Commissioning and model validation: test planning, witness testing and model validation against recorded performance.

Why Clients Choose Keentel

• Planning and design under one roof: the engineers who run the studies also write the specifications and review the designs.
• US standards fluency: daily work with NERC, IEEE, FERC and regional requirements.
• Independent advice: we are not tied to any equipment supplier, so recommendations follow the engineering.
• Practical, documented results: clear reports, auditable calculations and models that pass review.

Talk to an Engineer

Planning a new reactive compensation project, refurbishing an existing SVC or closing an interconnection reactive shortfall? Keentel Engineering can help you choose the right technology and size it correctly.

• Phone: (813) 389-7871
• Email: contact@keentelengineering.com
• Book a 15-minute call: calendly.com/keentel-engineering/15min

References

• IEEE Std 1031, IEEE Guide for the Functional Specification of Transmission Static Var Compensators.
• IEEE Std 1052, IEEE Guide for Specification of Transmission Static Synchronous Compensator (STATCOM) Systems.
• IEEE Std 519-2022, IEEE Standard for Harmonic Control in Electric Power Systems.
• IEEE Std 1453, IEEE Recommended Practice for the Analysis of Fluctuating Installations on Power Systems.
• IEEE Std 18, IEEE Standard for Shunt Power Capacitors; IEEE Std 1036, IEEE Guide for the Application of Shunt Power Capacitors.
• IEEE Std 693, IEEE Recommended Practice for Seismic Design of Substations.
• IEEE Std 2800-2022, IEEE Standard for Interconnection and Interoperability of Inverter-Based Resources Interconnecting with Associated Transmission Electric Power Systems.
• NERC Reliability Standards TPL-001-5.1, FAC-008, MOD-032, MOD-033, PRC-005, PRC-024, PRC-029 and CIP-002.
• FERC Order No. 827, Reactive Power Requirements for Non-Synchronous Generation (2016).
• N. G. Hingorani and L. Gyugyi, Understanding FACTS: Concepts and Technology of Flexible AC Transmission Systems, IEEE Press.
• R. M. Mathur and R. K. Varma, Thyristor-Based FACTS Controllers for Electrical Transmission Systems, IEEE Press/Wiley.
• Keentel Engineering, "STATCOM and Voltage Stability," companion technical blog.

Disclaimer

This article is for general technical information only. Worked examples use illustrative values and simplified assumptions; costs, losses and ratings are not quotations or design values. Standards are cited as of publication and may be revised; always confirm the current edition and its applicability with the relevant planning coordinator, transmission owner or regional entity. Project decisions require site-specific studies by qualified engineers.

© Keentel Engineering