If you are choosing between a dry type, a cast resin and an oil-immersed transformer, here is the short answer: cast resin wins indoor rooms that are humid, dirty or fire-restricted; VPI dry type wins clean, dry indoor rooms on budget; and oil-immersed wins outdoor pads, utility feeders and any site where a vault is not a problem.
But the short answer is not the whole story, because the cheapest unit to buy is rarely the cheapest unit to own. On a 25-year total cost of ownership (TCO) basis for a 1,000 kVA unit installed indoors, we calculate a VPI dry type at roughly USD 62,600 against USD 69,300 for oil-immersed — a gap of about USD 6,700 that comes almost entirely from vault construction and oil maintenance, not from the transformer itself. Move the same comparison outdoors to a pad, and the answer flips: oil-immersed becomes the cheaper option over 25 years.
This page compares all three constructions attribute by attribute, shows what the codes actually require, and then runs the numbers with a transparent 25-year model you can copy into your own RFQ.
| Your constraint | Recommended construction | Why |
|---|---|---|
| Indoor electrical room, basement, or public building | Cast resin | No oil, no vault; qualifies under the Class 155 insulation exception in NEC 450.21(B) when fully enclosed except for ventilation openings |
| Indoor room, clean and dry, capex-sensitive | VPI dry type | Same code position as cast resin at a lower purchase price; suitable where humidity and contamination are controlled |
| Outdoor pad, rural feeder, utility distribution | Oil-immersed | Lowest cost per kVA, smaller footprint, higher standard impulse level, no enclosure premium |
| Coastal, chemical or heavily polluted site | Cast resin (E2 / F1) | Encapsulated winding tolerates humidity and contamination that an open VPI winding cannot; stainless enclosure where specified |
| Maximum continuous overload headroom | Dry type (80 K rise design) | Up to 30% continuous overload without loss of life, plus up to 33% short-time reserve with forced air |
| Longest transformer-room-free schedule | Dry type | 2026 lead times run roughly 20–32 weeks for dry type against 40–65 weeks for pad-mount liquid-filled |
Before comparing, it helps to be precise about what each term means, because tenders frequently mix them up — and that leads to bids that cannot be compared.
A dry type transformer uses solid insulation and air — natural or forced — instead of a liquid dielectric. Inside this family there are two dominant constructions you will actually be asked to price:
VPI (vacuum pressure impregnated). The wound coil is dried under vacuum and then impregnated with varnish, so the winding is bonded but not fully encapsulated. Air circulates over the winding surface and heat leaves by convection. VPI units are lighter, easier to repair, lower in cost, and need a reasonably clean, dry air supply.
Cast resin. The winding is cast in epoxy resin under vacuum, so conductors are fully encapsulated in a solid dielectric. The casting protects the winding from moisture, dust and corrosive atmospheres, and gives the highest mechanical strength against short-circuit forces — which is why cast resin dominates indoor medium-voltage dry-type selection.
A useful rule of thumb: if the word “dry type” appears alone in a specification, ask which of the two is intended. The price difference between VPI and cast resin can exceed 15%, and the two are not interchangeable in a humid or contaminated room.
A cast resin transformer is a dry type transformer in which the core-coil assembly is encapsulated in cast epoxy resin. Under IEC 60076-11 it is normally specified with three class codes, and these three codes are where real projects are won or lost:
If the room is humid, coastal or unheated, insist on C2 / E2 / F1 and require the class codes to appear on the nameplate and the test certificate. Do not assume E2 simply because the datasheet says “dry type” — it is a design and test claim, not a category label. More detail on how VPI and cast resin differ in construction is covered separately, and cast resin benefits and limitations are set out in their own guide.
An oil-immersed transformer places the core-coil assembly in a sealed steel tank filled with mineral oil or a less-flammable ester fluid. Oil does two jobs at once: insulation and heat transfer. Because liquid convection removes heat far more effectively than air, the winding can run at a higher current density for the same temperature rise — which is why an oil-immersed unit at a given kVA is physically smaller and slightly lower in loss than a dry-type equivalent, and why the gap widens as ratings increase.
The trade-off is everything that comes with liquid: containment, fire-rated construction, oil testing, gasket maintenance and fluid disposal at end of life. For a broader treatment of the liquid-filled family, see our guide to what an oil-immersed transformer is and what it costs.
The table below compares the three constructions on the attributes that actually drive a purchasing decision. Ratings assume a three-phase distribution or small power transformer in the 500–2,500 kVA range.
| Attribute | VPI dry type | Cast resin dry type | Oil-immersed |
|---|---|---|---|
| Insulation / cooling medium | Solid insulation + air, varnish-impregnated winding | Solid insulation + air, epoxy-encapsulated winding | Mineral oil or ester fluid, sealed tank |
| Governing product standard | IEC 60076-11; IEEE C57.12.01; UL 1561 (≤600 V) or UL 1562 (>600 V) | IEC 60076-1 / -2; IEEE C57.12.00; IEEE C57.94 (maintenance) | |
| Typical insulation temperature rise classes | A / E / B / F / H — 60 / 75 / 80 / 100 / 125 K | Class A — top-oil 60 K, winding average 65 K | |
| Fire behaviour | Low fire load; no liquid | Class F1 self-extinguishing epoxy | Combustible mineral oil; ester fluids have a fire point ≥300 °C |
| Indoor code position (NEC Art. 450) | 450.21 — fire-resistant room above 112.5 kVA, unless the Class 155 exception applies | 450.26 — vault required for mineral oil indoors | |
| Short-circuit withstand | Good; winding bonded but not encapsulated | Highest — conductors locked in solid epoxy | Good; restrained by the tank and clamping structure |
| Sensitivity to humidity, dust, corrosive air | Highest — needs clean, dry air | Lowest — encapsulated winding | Low; tank is sealed, external parts still need coating specification |
| Footprint at equal kVA | Larger than liquid-filled; penalty grows with rating | Smallest of the three | |
| Weight and floor loading | Moderate, concentrated under the core | Moderate to heavy; resin mass adds to steel and copper | Heaviest overall (tank + oil), but compact footprint |
| Audible noise | Higher; enclosure or acoustic treatment may be needed | Higher; enclosure or acoustic treatment may be needed | Lower — the tank and oil damp core noise |
| Losses at equal kVA | Slightly higher than liquid-filled; the gap narrows as efficiency tier rises | Lowest of the three at equal efficiency tier | |
| Overload headroom | Up to +30% continuous with an 80 K rise design; +15% at 115 K; none at 150 K | +12% at a 55 K rise / 120 °C class; none at a 65 K rise | |
| Forced-air cooling reserve | Up to +33⅓% for substation-type dry units (225–10,000 kVA) | +15% at 501–2,000 kVA, +25% at 2,501–10,000 kVA; no fan cooling on pad-mount units | |
| Standard impulse (BIL) at 34.5 kV system | 150 kV per IEEE C57.12.01 (optional ratings available) | 150 or 200 kV per IEEE C57.12.00 | |
| Recurring maintenance | Cleaning, infrared thermography, insulation resistance or partial discharge test | Oil sampling, dielectric and dissolved gas analysis, gasket and bushing service | |
| End-of-life disposal | Straightforward; higher metal scrap credit | Fluid handling and disposal, possible PCB screening under 40 CFR 761 | |
| Typical 2026 lead time | About 20–32 weeks for a comparable class | About 40–65 weeks for pad-mount liquid-filled | |
Almost every comparison article on the internet makes the same mistake here. Under the US energy conservation standards in 10 CFR 431 Subpart K, the minimum efficiency tables for the three families are measured at different load points:
Putting a 35%-load figure next to a 50%-load figure and concluding that “dry type is 0.15% less efficient” is not a valid comparison. Load loss rises with the square of the load, so the two numbers are measuring different operating points. If you want a real answer, convert both to kW at your load factor — which is exactly what the TCO model further down this page does.
| Rating (three-phase) | LV dry type @ 35% load | MV dry type @ 50% load (20–45 kV BIL) | Liquid-immersed @ 50% load |
|---|---|---|---|
| 75 kVA | 98.60% | 98.33% | 99.03% |
| 150 kVA | 98.83% | 98.60% | 99.16% |
| 300 kVA | 99.02% | 98.82% | 99.27% |
| 500 kVA | 99.14% | 98.96% | 99.35% |
| 750 kVA | 99.23% | 99.07% | 99.40% |
| 1,000 kVA | 99.28% | 99.14% | 99.43% |
| 1,500 kVA | — | 99.22% | 99.48% |
| 2,500 kVA | — | 99.31% | 99.53% |
Two conclusions follow from these tables, and both are useful in a bid evaluation:
First, the efficiency gap between technologies is small; the gap between efficiency tiers is large. At 1,000 kVA the dry-type and liquid-immersed floors differ by roughly 0.15 percentage points. The difference between an entry-tier design and a premium low-loss design within the same family is typically several times larger. Efficiency is a tier decision, not a technology decision — so specify the tier, then compare technologies on TCO.
Second, environmental and harmonic derating hits dry type harder. Under IEEE C57.110, for the same six-pulse rectifier load spectrum, an air-cooled dry-type unit may only be usable to roughly 77% of nameplate against about 87% for a liquid-filled unit, because harmonic eddy losses in the windings are harder to remove through air than through oil. If the load is a UPS, a rectifier, a variable-frequency drive bank or a data-centre white space, include a K-factor requirement — not just a kVA rating — and ask the supplier to show the derating calculation. Our nameplate and ratings guide explains which of these values belong on the rating plate, and the site-wide insulation class and temperature rise reference covers the class codes in full.
The single largest financial difference between dry type and oil-immersed usually has nothing to do with the transformer. It is the concrete, drainage, fire rating and ventilation the installation code demands around it. NEC Article 450 Part II draws the line:
| Code clause | Applies to | Requirement |
|---|---|---|
| 450.21(A) | Dry type indoors, ≤112.5 kVA | At least 300 mm (12 in) separation from combustible material, unless separated by a fire-resistant heat-insulating barrier or fully enclosed except for ventilating openings (≤600 V) |
| 450.21(B) | Dry type indoors, >112.5 kVA | Fire-resistant transformer room, minimum 1-hour rating. Exception 1: Class 155 or higher insulation separated from combustibles by a barrier or ≥1.83 m horizontal and 3.7 m vertical. Exception 2: Class 155 or higher and completely enclosed except for ventilating openings |
| 450.21(C) | Dry type indoors, >35,000 V | Vault complying with Part III |
| 450.22 | Dry type outdoors | Weatherproof enclosure; above 112.5 kVA keep 300 mm from building combustibles unless Class 155 or higher and enclosed |
| 450.23 | Less-flammable liquid insulators indoors | Permitted where the listed conditions are met — the dividing line is a fluid fire point not below 300 °C, plus liquid confinement and, in the heavier path, automatic fire suppression |
| 450.26 | Oil-insulated indoors | Vault required (six exceptions for small or special units) |
| 450.27 | Oil-insulated outdoors | Safeguards against fire spread to buildings, fire escapes and openings — space separation, fire barriers, automatic suppression or oil-confining enclosures |
| Part III (450.41–450.48) | Vault construction | Walls and roof 3-hour fire resistance (reducible to 1 hour with sprinkler, water spray, CO₂ or halon protection); concrete floor ≥100 mm; 3-hour outward-swinging doors; outside-air ventilation; drainage; no foreign pipes or ducts; no storage |
Read 450.21(B) Exception 2 carefully, because it is the clause doing the real work on most data-centre and hospital projects: a dry-type unit above 112.5 kVA can avoid the fire-resistant room entirely if it carries a Class 155 or higher insulation system and is completely enclosed except for ventilating openings. That is a specification line item, not a default. If your tender does not state it, you may pay for a fire-rated room you did not need.
For oil-immersed units sited outdoors near buildings, separation distances are governed by loss-prevention guidance in addition to the code. As a reference point, an FM-approved liquid-insulated transformer is typically permitted at 3 ft from a wall, while a non-approved unit with under 500 gallons requires 5 ft to a 2-hour-rated wall, 15 ft to a non-combustible wall and 25 ft to a combustible wall. Concrete is not the cheap option — it is the option you pay for over 25 years.
These three attributes often decide a project after the price comparison has narrowed the field to two suppliers.
Continuous overload without loss of life. A dry-type unit built to an 80 K rise in a 220 °C insulation system can carry up to 30% above nameplate continuously; at a 115 K rise the figure is 15%; at 150 K there is no margin. A liquid-filled unit at a 55 K rise in a 120 °C class gives about 12%; at 65 K rise there is none. Note the direction of the trade: buying a lower temperature-rise design costs more up front and buys overload margin — and it buys more of it on the dry-type side, where the class structure runs higher.
Forced-air cooling reserve. If the load forecast is uncertain, the cheapest flexibility is often a fan kit rather than a larger transformer. Dry-type substation units from 225 to 10,000 kVA gain up to 33⅓% capacity with fans. Oil-filled substation units gain 15% between 501 and 2,000 kVA and 25% between 2,501 and 10,000 kVA, while pad-mount liquid-filled units typically have no fan cooling option at all. If growth is likely and the site is a pad, that limitation is worth knowing before you buy.
Surge withstand. Basic impulse level is where liquid-filled units keep a real advantage at medium voltage. At a 25 kV system voltage the standard dry-type BIL is 110 kV against 125 or 150 kV for a liquid distribution unit; at 34.5 kV the standard dry-type value is 150 kV against 150 or 200 kV. On exposed overhead-fed sites with high lightning exposure, that difference is a technical argument for oil, or for specifying a higher-than-standard BIL on the dry unit — an option that exists but costs money.
Maintenance is the second-largest ongoing cost difference after civil works, and the only one that is almost entirely predictable.
| Activity | VPI / cast resin dry type | Oil-immersed |
|---|---|---|
| Routine visual and thermal inspection | Annual — dust removal, infrared scan of terminations | Annual — tank, bushings, gaskets, oil level, silica gel |
| Dielectric condition monitoring | Insulation resistance (megger) and, on critical units, partial discharge testing | Oil sampling per ASTM D923 with dielectric strength, moisture and dissolved gas analysis |
| Consumables | None beyond cleaning materials | Oil, gaskets, seals, desiccant, occasional reconditioning or filtration |
| Typical spacing between major interventions | Insulation assessment on a 5–10 year cycle | Oil processing roughly every 5–8 years depending on results; gasket service as leakage appears |
| Design service life (IEEE C57.91 / C57.96) | Both are designed for 20–30 years, and both have been known to run beyond 40 years under normal loading and conditions. Neither standard can predict life precisely — loading history and thermal duty dominate. | |
| End of life | Resin and metal; higher scrap credit, no fluid handling | Fluid handling, possible PCB screening; documentation burden is higher |
| Failure modes to watch | Winding contamination and partial discharge in wet or dusty rooms | Gasket leaks, bushing failure, moisture ingress, oil degradation |
One practical point on dry type: the maintenance advantage only holds if the room is clean. A VPI winding in a dusty, humid room will accumulate contamination and develop partial discharge. Either specify cast resin for that room, or commit to the cleaning programme. There is no third option where the low-maintenance claim survives contact with a bad room.
For distribution transformers, the standard way to evaluate losses is capitalisation: assign a present value to every watt of loss, then add it to the purchase price. The method is set out in the European harmonisation documents HD 428 and HD 538, and it reduces to one line:
TCO = purchase price + (A × P₀) + (B × Pk)
where P₀ is rated no-load loss in watts, Pk is rated load loss in watts, and A and B are the capitalised cost per watt:
A = [(1 + i)n − 1] ÷ [i × (1 + i)n] × energy price × 8,760
B = A × (actual load ÷ rated load)²
The first term is the annuity factor for lifetime n years at discount rate i. The 8,760 figure is simply the hours in a year, because no-load loss is present whenever the transformer is energised — around the clock, whether the building is busy or empty. That is why no-load loss dominates the loss bill at low load factors, and why load loss takes over at high ones.
| Parameter | Value used on this page | Resulting capitalised cost |
|---|---|---|
| Discount rate (i) | 8% | Annuity factor 10.673 for 25 years |
| Service life (n) | 25 years | |
| Energy price | USD 0.12 / kWh | A = USD 11.22 per watt of no-load loss |
| Average loading — 35% | B = A × 0.1225 | USD 1.37 per watt of load loss |
| Average loading — 50% | B = A × 0.25 | USD 2.81 per watt of load loss |
| Average loading — 75% | B = A × 0.5625 | USD 6.31 per watt of load loss |
| Average loading — 100% | B = A | USD 11.22 per watt of load loss |
A very common procurement mistake is to ask for the losses after the supplier has been chosen. The HD 428 practice is the opposite: state A and B in the RFQ so that every bidder designs to the same loss evaluation. A bidder who knows that no-load loss is valued at USD 11.22 per watt will price a better core; a bidder who does not will price the cheapest core that passes the efficiency floor.
Assume a 1,000 kVA three-phase unit in an indoor electrical room of a commercial or light industrial building, 50% average loading, the energy price and discount rate above, and a 25-year life. Loss figures are indicative for this rating — substitute the guaranteed losses from your supplier’s test report before you make a decision.
| Cost block over 25 years | VPI dry type | Cast resin | Oil-immersed |
|---|---|---|---|
| 1. Transformer purchase price | 16,500 | 18,000 | 13,500 |
| 2. Freight, rigging, commissioning | 990 | 1,080 | 1,100 |
| 3. Civil works — fire-rated room and ventilation; vault, 3-hour walls, drainage, oil pit and barriers | 2,200 | 2,200 | 9,500 |
| 4. No-load loss energy, present value (A × 1,600 W / 1,600 W / 1,300 W) | 17,952 | 17,952 | 14,586 |
| 5. Load loss energy at 50% loading, present value (B × 7,600 W / 7,600 W / 6,200 W) | 21,356 | 21,356 | 17,422 |
| 6. Maintenance, testing and consumables | 3,000 | 3,000 | 11,000 |
| 7. Cooling auxiliary power | 0 | 0 | 0 |
| 8. End-of-life disposal, net of scrap value | 600 | 600 | 2,200 |
| 25-year total cost of ownership (USD) | 62,598 | 64,188 | 69,308 |
Three things are worth pulling out of this table:
The transformer itself is a minority of the cost. Purchase price is 22–26% of 25-year TCO here. Energy losses alone are worth more than the unit.
Oil-immersed is the cheapest to buy and the most expensive to own indoors. Its USD 3,000 purchase-price advantage is erased almost twice over by USD 7,300 of extra civil works and USD 8,000 of extra maintenance. The losses are actually in oil’s favour by about USD 7,300 over 25 years — the vault and the oil programme are what decide the outcome.
VPI and cast resin are close, so decide on the room. The USD 1,590 gap between them is a construction and environment decision, not a loss decision. If the room is clean and dry, VPI is the better buy. If there is condensation, dust or corrosive air, the cast resin premium is insurance you should take — and the TCO difference is under 3%.
Now move the unit to an outdoor pad beside the building, raise average loading to 60%, and keep everything else the same. Civil works collapse to a pad plus spill containment and separation distances, and the oil programme still applies.
| Cost block over 25 years | VPI dry type (NEMA 3R) | Cast resin (E2 / F1) | Oil-immersed (pad) |
|---|---|---|---|
| 1. Transformer purchase price | 16,500 | 18,000 | 13,500 |
| 2. Weatherproof enclosure premium | 1,200 | 1,200 | included |
| 3. Freight, rigging, commissioning | 990 | 1,080 | 1,100 |
| 4. Civil works — pad, spill containment, separation | 800 | 800 | 1,500 |
| 5. No-load loss energy, present value | 17,952 | 17,952 | 14,586 |
| 6. Load loss energy at 60% loading, present value (B = 4.04/W) | 30,704 | 30,704 | 25,048 |
| 7. Maintenance, testing and consumables | 3,000 | 3,000 | 11,000 |
| 8. End-of-life disposal, net of scrap value | 600 | 600 | 2,200 |
| 25-year total cost of ownership (USD) | 71,746 | 73,336 | 68,934 |
The ranking has inverted. Outdoors, oil-immersed is now the lowest-TCO option by about USD 2,800 against VPI, because the civil penalty that dominated the indoor case has almost disappeared. This is the whole argument in one pair of tables: the transformer did not change, the site did.
Two variables move the indoor comparison more than anything else: how hard the unit is loaded, and what electricity costs. Holding civil works, maintenance and purchase price at the indoor levels above, the table below shows the 25-year cost difference between oil-immersed and a VPI dry type. A positive number means oil-immersed costs more over 25 years; a negative number means it is the cheaper choice.
| Energy price | 35% loading | 50% loading | 75% loading | 100% loading |
|---|---|---|---|---|
| USD 0.06 / kWh | +11,365 | +10,363 | +7,908 | +4,473 |
| USD 0.12 / kWh | +8,720 | +6,710 | +1,810 | −5,064 |
| USD 0.20 / kWh | +5,193 | +1,855 | −6,326 | −17,780 |
Three practical readings:
Two further variables are worth noting. Higher discount rates compress future energy savings and favour the cheaper unit up front; rising tariffs do the opposite. And any load profile that is idle overnight massively increases the weight of no-load loss, because no-load loss is capitalised over 8,760 hours a year whether or not there is any load — a point that matters for solar and storage projects especially.
| Application | Typical choice | Deciding factor |
|---|---|---|
| Data centre white space | Cast resin | Fire code and indoor siting; K-factor for harmonic load; N+1 redundancy changes the loading profile |
| Hospitals and healthcare | Cast resin | Fire safety, indoor siting, low maintenance in restricted areas |
| Commercial building main substation | Cast resin or VPI | Room quality and budget; either qualifies under the 450.21(B) exception if Class 155 or higher and enclosed |
| Industrial plant, outdoor substation | Oil-immersed | Cost per kVA, overload serviceability, harsh ambient |
| Metro, tunnel, confined space | Cast resin | No flammable liquid in an occupied or hard-to-access space |
| Coastal or chemical plant | Cast resin with E2 / F1 classes | Encapsulated winding and stainless enclosure where specified |
| Utility distribution, pole or pad | Oil-immersed | Cost per kVA and standard lead-time class |
| Solar and storage collector substation | Oil-immersed outdoors, dry indoors | Long idle hours make no-load loss the dominant loss; PCS harmonics require a K-factor check either way |
| Mining and heavy industry | Oil-immersed outdoors, cast resin underground | Dust, vibration and thermal cycling outside; no flammable liquid where access is restricted |
Most comparison mistakes are specification mistakes. These are the items that let you evaluate bids on equal terms — and the ones that let you run the TCO model above with real numbers instead of estimates.
| Require from every bidder | Why it matters |
|---|---|
| Guaranteed no-load loss P₀ and load loss Pk in watts, with the reference temperature stated | Without both figures you cannot capitalise losses, and you cannot compare a dry-type quote to a liquid-filled quote |
| Efficiency tier and the standard it is claimed against | Tier, not technology, is the main efficiency lever |
| Insulation system class and temperature rise class | Determines overload headroom and eligibility for the NEC 450.21(B) exception |
| IEC 60076-11 class codes C / E / F for dry units | E2 and F1 are the difference between surviving a humid room and failing in one |
| Forced-air cooling reserve, and the kVA it unlocks | Often cheaper growth capacity than buying a larger transformer now |
| Impedance, vector group, tap range and BIL | Needed for protection coordination and for surge-exposed sites |
| K-factor or derating calculation for harmonic loads | IEEE C57.110 derating differs materially between air-cooled and liquid-filled units |
| Enclosure rating, finish standard and material | Drives coastal and outdoor service life; specify stainless where corrosion applies |
| Routine, type and special test reports, and whether a witness test is required | Test data is the evidence behind every loss guarantee above |
| Site conditions — ambient, altitude, humidity and indoor or outdoor | All three change the rating and the derating calculation |
| Delivery schedule and shipping basis | With 2026 lead times running 20–32 weeks for dry type against 40–65 weeks for pad-mount liquid-filled, schedule is a real cost |
| Loss evaluation factors A and B, stated by the buyer | The HD 428 practice: when every bidder prices the same loss valuation, the bids become comparable |
One standards trap worth knowing: UL 1561 covers dry-type general-purpose and power transformers rated 600 V and below, while UL 1562 covers dry-type distribution transformers above 600 V. Specifications that pair “cast coil” with “UL 1562” on a 480 V unit are describing a listing that does not exist for that product — and those bids will not be comparable. Voltage class sets the standard; check it before the tender goes out.
Further reading on this site: dry type transformer specifications and ratings, dry type transformer price guide, power transformer price per kVA, how to choose a dry type transformer supplier, evaluating an oil-immersed distribution transformer manufacturer, and the full power transformer selection guide.
No — cast resin is a type of dry type transformer. Dry type is the wider family of transformers that use solid insulation and air instead of liquid. Within it, VPI (vacuum pressure impregnated) and cast resin are the two constructions you will be asked to price. The distinction matters commercially: cast resin typically carries a premium of roughly 10–15% over an equivalent VPI unit, and it buys you an encapsulated winding that tolerates humidity, dust and corrosive air.
Choose cast resin when the unit goes indoors in a room that is humid, dusty, coastal or unheated; when the site is fire-restricted or the presence of oil is unacceptable; when short-circuit forces are high and you want the maximum mechanical strength; or when the specification calls for C2 / E2 / F1 class codes. For a clean, dry, air-conditioned electrical room on a tight budget, VPI is usually the better value.
On purchase price, yes — typically 20–30% more at the same rating. On 25-year total cost of ownership the answer depends on the site. Indoors, the dry-type premium is usually recovered because an oil-immersed unit needs a vault with 3-hour fire-rated construction, drainage and ventilation, plus a continuing oil testing and maintenance programme. Outdoors on a pad, where civil works are minor, oil-immersed usually wins on TCO — and at high load factors or high electricity prices its lower losses make it the cheaper choice even indoors.
Less than liquid-filled units, but not none. Plan on an annual inspection with dust removal and infrared thermography of the terminations, plus insulation resistance measurement and, on critical units, partial discharge testing. The low-maintenance reputation only holds in a clean, dry room. In a dusty or humid room a VPI winding will contaminate and develop partial discharge — in that environment, specify cast resin rather than relying on the maintenance programme to compensate.
No, and this is the most common error in transformer bid evaluation. Under 10 CFR 431 Subpart K the efficiency floor for low-voltage dry-type transformers is stated at 35% of nameplate load, while the floors for medium-voltage dry-type and liquid-immersed transformers are stated at 50% of nameplate load. Load loss rises with the square of the load, so those two numbers describe different operating points. Convert both datasheets to total loss in kW at your own average load factor before comparing them.
Both families are designed against a 20–30 year life, and both have run beyond 40 years in service under normal conditions. IEEE C57.91 for liquid-immersed and IEEE C57.96 for dry type both state that life cannot be predicted with precision. In practice loading history and thermal duty dominate: a unit run continuously at high temperature will age faster than its nameplate suggests, whichever construction you buy. That is another reason to specify a lower temperature-rise design where overload margin is expected.
Start with the site, not the transformer. If the unit has to sit indoors in a fire-restricted, humid or contaminated room, cast resin is the answer and the modest premium over VPI is well spent. If the room is clean and dry, VPI gives you the same code position for less money. If the unit can live outdoors on a pad, oil-immersed will usually be the lowest-cost choice over its life — unless you are indoors, lightly loaded and paying high tariffs, in which case the vault penalty outweighs the loss advantage and dry type wins. Then confirm the decision with the 25-year model, using loss figures from the supplier’s test report and loss evaluation factors A and B written into the RFQ.
Kampa Electric manufactures all three constructions — VPI and cast resin dry type transformers through our dry-type transformer insulation series, and oil-immersed units from our oil-immersed distribution transformer range. If you send us the site conditions, load profile, tariff and discount rate, we will return a loss-evaluated comparison of the three options built on your numbers. Contact our engineering team or review our transformer manufacturing capabilities first.
