Designing Substation Transformers for Solar Farms With High DC-to-AC Loads

At the point where a solar farm moves from inverter blocks to the collector system, transformer selection can become a source of avoidable risk. A nameplate voltage match may look acceptable, yet the unit may still face uneven daytime loading, harmonic heating, repeated voltage changes, reactive-power swings, and high ambient temperature. The practical answer is to design the substation transformer around the inverter-side AC duty cycle and grid interconnection requirements—not around the DC capacity alone.

A transformer does not normally carry the photovoltaic array’s DC output directly. DC is converted by inverters before it reaches the medium-voltage collection network. However, a high DC-to-AC ratio changes the operating profile seen by the transformer. It can create long periods near the AC export limit, sharper ramps around changing irradiance, clipping-related full-load operation, and more frequent inverter control activity. A properly designed substation transformer for solar farms must therefore be evaluated for thermal performance, losses, impedance, harmonics, insulation duty, protection coordination, and the exact grid-code functions required at the point of connection.

Start with the actual AC loading profile

The DC-to-AC ratio is often used to improve energy harvest during lower-irradiance periods. For example, PV module capacity may exceed the aggregate inverter AC rating. That arrangement does not automatically require a transformer sized to the full DC nameplate. The relevant electrical limit is usually the maximum simultaneous AC output from the inverter system, adjusted for auxiliary demand, reactive-power operation, contingency operating modes, and any contractual export limit.

The difficulty is that “maximum AC output” is not always a single stable number. Inverter controls may be required to supply or absorb reactive power while active generation remains high. Depending on the grid operator’s requirements, the transformer may need to support an apparent-power duty higher than the expected MW export alone. A design based only on active power can understate current and thermal loading.

Before specifying ratings, establish a time-based operating envelope rather than relying on one annual average. The engineering review should define:

• Maximum active power delivered from all inverters at the transformer terminals.
• Required reactive-power range at high active-power output.
• Expected duration of clipping or operation near the AC export ceiling.
• Ambient temperature during the highest expected electrical loading period.
• Auxiliary loads, station-service demand, battery systems, or shared feeder loads.
• Whether one transformer may temporarily carry load from an adjacent inverter block after a planned or unplanned outage.

This approach distinguishes an installation with a high DC-to-AC ratio but modest transformer duty from one that will spend long summer periods close to its thermal limit. It also prevents an oversimplified conclusion that more DC capacity automatically requires a proportionally larger transformer.

Thermal design must reflect solar operating conditions

Solar generation is cyclic, but that does not make transformer cooling easy. Peak generation commonly occurs when outdoor temperature, solar radiation on enclosures, and local equipment-room temperature are also elevated. In a compact substation, heat rejected by inverters, switchgear, and cable systems can raise the air temperature around a dry-type transformer. Dust accumulation, restricted ventilation paths, or poorly arranged louvers can further reduce cooling performance.

Thermal evaluation should consider the hottest credible operating combination: high inverter output, reactive-power demand, high ambient temperature, and reduced airflow. A transformer that operates comfortably at rated load in a controlled factory test environment may require derating in a solar plant enclosure or containerized skid.

For liquid-filled transformers, the review focuses on top-oil temperature, winding hot-spot temperature, radiator exposure, cooling stages, and oil expansion arrangements. For cast-resin and other dry-type designs, winding temperature rise, ventilation volume, enclosure configuration, fan control, and temperature-monitoring points become especially important. Neither technology is universally better; the site layout, fire-safety requirements, environmental conditions, maintenance strategy, and project standards determine the appropriate choice.

Forced-air cooling can be useful when it is treated as a defined operating mode rather than an assumed permanent condition. The fan supply, automatic controls, alarms, maintenance access, and fallback rating without forced air should all be documented. A transformer with an overload capability under forced-air cooling should not be used to compensate for a fundamentally undersized base rating unless the expected duty cycle, thermal model, and protection settings support that decision.

Harmonics: assess the waveform, not just the fundamental current

Modern solar inverters are designed to deliver controlled AC output, but their switching operation can introduce harmonic and high-frequency components. The actual level at the transformer depends on inverter topology, filter design, operating point, cable lengths, grounding arrangement, network impedance, and interactions among parallel inverter blocks. Harmonic current increases conductor losses and can produce additional stray losses in windings, clamps, tank walls, and structural parts.

Do not assume that a compliant inverter data sheet alone completes the transformer harmonic assessment. The evaluation should review the inverter harmonic spectrum at relevant load levels, the number of units connected in parallel, collector-cable capacitance, and the impedance of the upstream grid. Harmonic resonance can emerge from the combination of equipment even when each item is acceptable in isolation.

Questions that should be resolved during design include whether the transformer manufacturer needs harmonic-loss information, whether a derating calculation is required, and whether the specified impedance will affect resonance or filter performance. In larger collector systems, a network harmonic study may be necessary to determine whether filters, modified transformer design, or operational limits are needed.

High-frequency common-mode effects also deserve attention at the inverter-transformer interface. Cable shielding, grounding, surge-protection coordination, and winding insulation design should be considered together. These issues are particularly relevant where long MV cable runs connect several inverter stations or where the transformer is installed close to inverter outputs with fast switching edges.

Impedance is a grid-performance decision

Transformer impedance is often treated as a catalog value, but in a solar installation it has direct consequences for voltage regulation, fault current, inverter fault-ride-through behavior, and protection sensitivity. Excessively high impedance can increase voltage rise during export and reduce fault current available to protective devices. Very low impedance may increase short-circuit duty on switchgear and downstream equipment.

The preferred value depends on the collector layout and the strength of the grid connection. A short, stiff feeder presents a different condition from a long radial collector circuit with multiple inverter transformers. The study should examine voltage at the inverter terminals during full export, low-export conditions, reactive-power commands, and credible switching events. It should also test whether the inverter voltage-control functions remain stable with the selected transformer impedance and cable impedance.

Tap selection is equally important. Solar plants may experience a persistent voltage rise at the collector bus during high export, while the upstream voltage can vary with grid conditions. Fixed off-circuit taps may be adequate where the system study confirms a narrow voltage range. Where voltage variation is broader, the design may need a different regulation strategy, such as on-load tap changing at the main grid transformer or coordinated inverter reactive-power control. The transformer itself should not be expected to solve a system-level voltage-control problem without confirming the wider network behavior.

Choose insulation and protection for the installation, not only the rating

Solar sites frequently place transformers outdoors, in prefabricated substations, or near inverter skids. The installation can expose equipment to dust, humidity, salt-laden air, condensation, ultraviolet exposure, sand, and seasonal temperature variation. These conditions affect bushing selection, enclosure class, terminal clearances, coating systems, cable sealing, and the suitability of dry-type cooling paths.

For an indoor or enclosed medium-voltage distribution point where fire performance and reduced liquid-fire risk are priorities, a cast-resin dry-type transformer may be appropriate. The 20kV Three-Phase Cast Resin Dry-Type Distribution Transformer is available with a 20 kV input and 0.4 kV or 0.415 kV output configuration, capacities from 50 kVA to 5000 kVA, and a high-voltage tap range of ±2 × 2.5%. Its stated IP23 protection level, F1 flame-retardant class, temperature protection and control system, and forced-air overload capability can be relevant where the arrangement calls for a dry-type distribution transformer. The exact enclosure, ventilation arrangement, voltage ratio, and capacity still need to be matched to the solar block’s AC-side design.

Protection design should account for a key characteristic of inverter-based generation: inverter fault current is generally limited and may be controlled differently from the current supplied by rotating generators. Conventional overcurrent protection settings that work in a utility or industrial plant may not be sufficiently sensitive in a solar collector system. Differential protection, restricted earth-fault protection, feeder protection, and backup overcurrent elements should be coordinated with inverter fault-current behavior and transformer inrush.

Transformer energization deserves separate attention. Inrush current can cause nuisance trips, especially where several transformers are energized after a site-wide outage. The protection philosophy should distinguish inrush from internal fault conditions while remaining sensitive to genuine winding or terminal faults. Transformer vector group, grounding method, neutral treatment, and any zero-sequence paths must be coordinated with the collector system’s earth-fault scheme.

A practical design review before releasing the specification

A useful review sequence begins by drawing the power path from inverter terminals through the LV or MV collection system to the point of interconnection. Identify every location where power factor, voltage, current, or fault level changes. Then compare the transformer specification against the following engineering evidence rather than treating it as a standalone procurement document.

1. Confirm the rating basis. Use maximum inverter AC output and required kvar support to determine the continuous apparent-power duty. State whether the rating assumes natural cooling, forced cooling, or a limited overload period.
2. Check the voltage ratio and tap range. Verify normal, minimum, and maximum voltage at both transformer terminals under export and reactive-power conditions. Include cable voltage drop or rise.
3. Review impedance with the system model. Test fault duty, voltage regulation, protection reach, and the impact on inverter controls before accepting a nominal impedance value.
4. Provide harmonic input data. Give the manufacturer inverter harmonic information where available, and determine whether additional winding or stray-loss allowance is required.
5. Define the physical environment. State indoor/outdoor location, maximum ambient temperature, altitude, solar exposure, pollution level, moisture conditions, and ventilation restrictions.
6. Coordinate protection and monitoring. Confirm current-transformer ratios, temperature alarms, trip thresholds, fan alarms, surge arresters, earth-fault philosophy, and communications interfaces.
7. Plan maintainability. Confirm access for inspections, infrared scanning, cleaning, cable testing, fan replacement where applicable, and safe isolation.

Where designs most often lose margin

One frequent error is selecting a transformer based on PV DC capacity and then adding excessive margin without checking the actual inverter AC limitation. This can increase capital cost, no-load losses, footprint, and short-circuit duty without improving the real operating constraint. The opposite error is more serious: choosing a unit equal to nominal MW export while overlooking reactive-power demand, high ambient temperature, and transformer losses. In that case, the nameplate can appear adequate while winding temperature margin is insufficient.

Another weak approach is to specify “solar-duty” equipment without defining the duty. That phrase has little value unless it is translated into measurable requirements: load cycle, ambient conditions, harmonic content, cooling mode, insulation class, tap arrangement, impedance tolerance, monitoring devices, and required standards. Clear input data allows the manufacturer to design and test the appropriate equipment; vague language often leads to assumptions that surface only during commissioning.

Technical evaluators should also distinguish between a block transformer and the main substation transformer. A block unit near inverters may prioritize compactness, fire performance, low-voltage winding arrangement, and local thermal conditions. The main step-up transformer must usually address higher voltage insulation coordination, grid fault levels, system grounding, voltage regulation, and utility protection interfaces. Applying one set of assumptions to both locations can create gaps in the design.

When further system studies are justified

Additional analysis is warranted when the site has a weak grid connection, a long collector network, substantial cable capacitance, multiple inverter technologies, energy storage connected to the same bus, or strict reactive-power and fault-ride-through requirements. In these cases, load flow, short-circuit, harmonic, insulation-coordination, and protection studies are not paperwork additions; they determine whether the transformer parameters support stable operation.

The most reliable specification is one that states the electrical operating envelope first and the transformer construction second. When high DC-to-AC loading is translated into realistic AC current, kvar demand, temperature exposure, and inverter interaction, the transformer can be sized and configured for dependable service rather than merely selected by voltage and kVA.