Sumpner's Test in Transformer: What It Is, How It Works, and Why It Still Matters

Sumpner's test, widely known as the back to back test, is a method used to determine the efficiency, voltage regulation, and temperature rise of a transformer under conditions that closely mirror actual full load operation. Developed by British engineer William Edward Sumpner in 1891, the test was created to solve a practical problem in early transformer manufacturing: how do you test a large power transformer at full load without either wasting enormous amounts of electricity or requiring an equally large external load bank. The technique remains in use today, more than a century later, because the underlying problem it solves has never gone away.

This method is distinct from the open circuit test and the short circuit test, both of which are also used to evaluate transformer performance. Those two tests are simpler and require only one transformer, but they each reveal only one type of loss at a time. Sumpner's test uses two identical transformers connected together so that both iron loss and copper loss occur simultaneously, exactly as they would in a transformer running under real load. This makes the results more representative of actual field performance, particularly for temperature rise studies.

This guide walks through why this test exists, how the circuit is built, what each instrument reading means, how to calculate efficiency and regulation from the results, and where this concept shows up in Indian competitive exams and industry practice.

Why Do We Need Sumpner's Test When OC and SC Tests Already Exist

Picture a large power transformer built for a substation, something rated at several thousand kVA. To test it at full load the traditional way, you would need a load bank capable of absorbing that entire rated power, plus you would burn through that same amount of electricity for every hour of testing. For a small distribution transformer this is manageable. For a large power transformer, it becomes expensive, impractical, and in many cases simply impossible to arrange.

This is exactly why the open circuit test and short circuit test were developed first. The open circuit test measures iron loss by exciting the transformer at rated voltage with the secondary left open, so no load current flows and only core losses are recorded. The short circuit test measures copper loss by applying a small voltage to circulate rated current through the windings while the secondary is shorted, so only winding losses show up. Together, these two tests give you the equivalent circuit parameters of the transformer, and from those parameters you can calculate efficiency and regulation on paper.

There is a catch, though. In a transformer that is actually running under load, iron loss and copper loss happen together, all the time, continuously heating the core and the windings at once. The OC and SC tests never let you observe that combined effect because each test isolates only one loss. You cannot use either test to find out how hot the transformer actually gets after running at full load for several hours, because neither test puts real heat-generating current through both the core and the windings simultaneously. That gap is precisely what Sumpner's test was designed to close, and it is also the biggest technical differentiator between Sumpner's test and the OC/SC pair, something that becomes clearer once you see the circuit itself.

The Working Principle Behind Sumpner's Test

The genius of Sumpner's test lies in a clever electrical trick: connect two identical transformers so that one effectively "loads" the other, without either of them ever being connected to an external load bank. Think of it like two people passing a heavy box back and forth between themselves in a closed loop. Energy circulates between the two transformers, doing real work and generating real heat, while the wall socket only needs to supply enough power to cover the losses, not the entire circulating power.

To make this possible, the test requires two transformers with identical ratings, identical turns ratio, and matching impedance characteristics. This is the one hard requirement of the test, and it is also its biggest practical limitation, because manufacturers do not always have two identical units available for testing at the same time.

Here is how the connection is arranged. The primary windings of both transformers are connected in parallel and supplied with rated voltage at rated frequency from the main AC supply. The secondary windings are connected in series, but in phase opposition, meaning the induced electromotive forces of the two secondaries cancel each other out. When this cancellation is correct, the net voltage in the secondary loop reads zero, and no current circulates in that loop under this condition alone.

At this stage, since no current flows in the secondary loop, each transformer behaves exactly as if it were undergoing an open circuit test. The wattmeter connected on the primary side, usually labelled W1, therefore reads the combined iron loss of both transformers put together. Since the two transformers are identical, dividing this reading by two gives you the iron loss of a single unit.

The next step introduces a small additional voltage into the secondary loop using a low voltage regulating transformer, which is also powered from the same main supply. This injected voltage is gradually increased until the ammeter in the secondary loop shows the rated full load current. Because the secondaries are in phase opposition, this circulating current also causes rated full load current to flow through both primary windings. At this point, both transformers are electrically experiencing full load current in exactly the way they would during real operation, even though there is no external load anywhere in the circuit.

A second wattmeter, W2, placed on the regulating transformer side, now reads the combined full load copper loss of both transformers. This is essentially a short circuit test condition being created for both machines simultaneously, using only the small voltage needed to push full load current around the loop, not the full rated voltage. Divide the W2 reading by two, and you get the copper loss of a single transformer at full load.

Because W1 supplies the core losses and W2 supplies the copper losses of both machines, and both losses occur at once, the transformers heat up exactly as they would under continuous full load service. This is why Sumpner's test is also called the heat run test in some references, and why it is considered the most realistic way to check temperature rise. The setup is typically run continuously for 36 to 48 hours, with oil or winding temperature recorded at regular intervals, so engineers can see how the temperature stabilizes over time.

Circuit Diagram and Instrument Setup Explained Simply

To visualize the full circuit, imagine two identical transformers, T1 and T2, sitting side by side on a test bench.

On the primary side, both transformers draw power from the same AC mains supply. A voltmeter, an ammeter, and wattmeter W1 sit between the mains and the parallel-connected primaries. This tells you the supply voltage, the no-load current drawn by both transformers combined, and the combined iron loss.

On the secondary side, the two windings are joined in series opposition. A second voltmeter is placed across this loop first, purely to verify polarity. If the two secondaries are correctly connected in phase opposition, this voltmeter should read close to zero. If instead it reads close to double the rated secondary voltage, the connection is wrong, and the leads need to be swapped before proceeding, because a wrong connection here would cause a short circuit the moment current starts flowing.

Once the polarity check passes, a small regulating transformer, along with an ammeter and wattmeter W2, is inserted into this secondary loop. The regulating transformer injects a small adjustable voltage, the ammeter confirms when rated full load current is flowing, and W2 records the copper loss that results.

This entire arrangement is why the test is called "back to back": one transformer's secondary output effectively feeds back into the other transformer's secondary, creating a closed energy loop that only needs a trickle of power from the mains to sustain, even though full load current flows throughout.

Calculating Efficiency and Voltage Regulation from Test Results

Once the readings from W1 and W2 are recorded, calculating performance figures for a single transformer is straightforward, since the setup uses two identical units.

Iron loss per transformer (Pi): Divide the W1 reading by 2.

Full load copper loss per transformer (Pcu): Divide the W2 reading by 2.

Total loss per transformer at full load: Add Pi and Pcu together.

Efficiency at full load, expressed as a percentage, is calculated using the standard transformer efficiency formula:

Efficiency = (Output / (Output + Pi + Pcu)) x 100

Here, output refers to the rated kVA output of the transformer multiplied by the power factor of the load it is meant to serve. Since Pi and Pcu were measured directly under conditions that mimic real full load operation, this efficiency figure is considered more trustworthy than one calculated purely from OC and SC test parameters on paper.

Voltage regulation can also be estimated from the readings, since the equivalent resistance and reactance of the transformer can be derived from the voltage, current, and power values recorded during the copper loss stage of the test. A lower regulation value indicates the transformer maintains its output voltage more steadily as load is applied, which is a desirable trait for distribution transformers supplying sensitive loads.

Worked example: Suppose two identical 20 kVA transformers are tested using Sumpner's method. W1 reads 200 W and W2 reads 320 W at rated full load current. Iron loss per transformer works out to 100 W, and full load copper loss per transformer works out to 160 W. At a load power factor of 0.8 lagging, the output per transformer is 20,000 x 0.8 = 16,000 W. Efficiency then comes out to (16,000 / (16,000 + 100 + 160)) x 100, which is approximately 98.4 percent. This kind of quick calculation is exactly the type of numerical question that shows up in GATE and SSC JE papers.

Where This Concept Matters in Indian Engineering Education and Careers

Sumpner's test is a recurring, high-value topic in Indian electrical engineering curricula and competitive exams. It appears regularly in GATE Electrical Engineering, SSC JE Electrical, and RRB JE syllabi under transformer testing and machine performance sections, often as both a conceptual question and a numerical problem involving efficiency or regulation calculations from given wattmeter readings. Questions frequently test whether students understand why OC and SC tests cannot replace this method for heat run studies, so understanding the reasoning behind the test, not just memorizing the circuit, pays off directly in exam scoring.

Beyond exams, this topic connects directly to real testing work carried out in India's transformer manufacturing and power sector. Companies such as BHEL, Siemens, ABB, Crompton Greaves, and Telawne Power Equipments, along with state power utilities and testing laboratories accredited by the Central Power Research Institute, routinely conduct heat run and load testing on transformers before they are commissioned into the grid. Roles such as transformer testing engineer, quality test engineer, and graduate engineer trainee in transformer manufacturing units frequently list transformer testing knowledge, including back to back and heat run testing, as a core requirement. Fresh electrical engineering graduates entering testing, commissioning, or quality roles in this space typically start in the range of 3 to 6 LPA, with experienced testing engineers and specialists in transformer diagnostics earning considerably more as they take on senior technical or plant-level responsibilities.

For students aiming at government sector roles, a solid grasp of transformer testing methods also supports SSC JE and RRB JE preparation directly, since the electrical engineering syllabus for these exams explicitly covers transformer testing under its machines and apparatus sections.

Conclusion

Sumpner's test solves a genuine engineering problem: how to evaluate a transformer's real full load performance, including heating effects, without the cost and impracticality of an actual full load setup. By connecting two identical transformers back to back, with primaries in parallel and secondaries in phase opposition, the test recreates true full load conditions for both core and copper losses at once, while drawing only a fraction of the actual load power from the supply. The two wattmeter readings, once halved, give you iron loss and copper loss per transformer, which in turn let you calculate efficiency and voltage regulation with confidence. For students preparing for GATE, SSC JE, or RRB JE, and for anyone stepping into a transformer testing or quality role in India's power sector, understanding this test thoroughly, not just its circuit but the reasoning behind it, is time well spent.

Learn Sumpner's test in transformers, including its working principle, circuit, calculations, advantages, efficiency, and real-world applications.