ESP controllers — automatic voltage control

Spark-rate control. A spark is not a failure; it is the measurement. The controller deliberately operates at a set sparking frequency, because the voltage that produces occasional sparks is the highest voltage the gap will hold. On each spark the controller quenches the arc by cutting thyristor firing for a few half-cycles, drops the setpoint by a set step, then ramps back at a set rate. The setpoint spark rate, quench time, setback and ramp are all field-adjustable per bus section CONFIRM: Arrow standard spark-rate setpoint range and setback/ramp defaults.

Back-corona detection. On high-resistivity dust — rice-husk ash at 85–90 % amorphous silica is the classic case — voltage drop across the dust layer triggers a reverse discharge inside the layer itself. Current climbs while the useful field collapses: more milliamps, worse collection. A digital controller detects this from the slope of the operating V–I curve. In healthy corona, dV/dI is positive; as back-corona develops the slope flattens and then inverts. When the controller sees the characteristic slope change it stops chasing current and either reduces the operating point or switches the field to intermittent energisation.

How much ESP power can intermittent energisation save?

Intermittent energisation (IE) typically cuts corona power by 40–70 % at equal opacity. Instead of firing the thyristors every half-cycle, the controller energises a set number of half-cycles and skips the rest; on high-resistivity dust the skipped cycles suppress back-corona, so emissions hold or improve while power falls.

The mechanism is charge persistence: already-charged dust keeps migrating toward the plates during the de-energised half-cycles, while the pause lets charge bleed out of the layer before back-corona strikes. The charge ratio is selected per field CONFIRM: available IE charge-ratio range on Arrow-supplied controllers, the inlet field usually run harder than the outlet field.

On the design-basis 60 t/h bagasse-fired boiler, the four-field ESP absorbs 177 kW. A 40–70 % IE saving is 71–124 kW of continuous electrical load recovered — auxiliary power returned to export — while the train continues to meet its 24 mg/Nm³ design-basis outlet. Actual savings depend on resistivity and are confirmed during commissioning trials at stated load and opacity.

Analogue AVCs from the 1980s and 1990s hold a fixed setback and ramp, cannot detect back-corona, and offer no IE and no data link. The upgrade path is straightforward because the heavy electrical plant is retained: the TR set, HT cable and bus sections stay, the analogue chassis comes out, and a digital controller goes in on the same divider and shunt signals — converted one field at a time, so the ESP never fully de-energises. TR sets are insulation-tested before reuse; a set that fails becomes a transformer-rectifier sets replacement scope instead.

What can a controller retrofit fix — and what can it not?

A controller retrofit recovers whatever margin poor voltage control was wasting: slow ramps, excessive setbacks, undetected back-corona, no IE. It cannot fix gas velocity maldistribution, dust resistivity, worn electrodes or an undersized collecting area — a field held perfectly at spark-over still collects badly if the gas rushes through one corner of it.

The distinction decides retrofit economics. If a velocity traverse shows spread beyond σ ≤ 15 % of mean velocity, the money belongs first in gas distribution screens; if the plates are buckled or the emitters eroded, in the internals. Arrow surveys the precipitator — traverse, V–I curves per field, air-load test — before quoting, and states expected outlet on a defined basis (mg/Nm³, reference O₂, dry, load range). Guaranteed figures are stated per project after that technical assessment.

Arrow Energy Co., Ltd. supplies ESP controllers and complete control retrofits for precipitators built by other OEMs, as part of its ESP upgrade scope. Compatibility is electrical as much as mechanical: what must match is the TR nameplate rating, divider ratio, shunt rating and thyristor sizing, so every retrofit starts with a dimensional and electrical survey of the existing panels and TR sets. The replacement controller is configured and bench-tested against the surveyed values before shipment, then commissioned field by field against recorded air-load V–I curves.