Heat Exchanger Energy Balance: Comparing Hot-Side and Cold-Side Duties

Every heat exchanger calculation begins with a single number: the duty — the rate at which heat must pass from the hot fluid to the cold one, measured in kilowatts. Because the two fluids are separated by a thin wall and never mix, the duty can be calculated independently from either stream, and in steady operation the hot-side and cold-side results must agree. That agreement is the cheapest and most revealing check available on heat exchanger data: it exposes wrong flows, wrong specific heats, and unaccounted losses on paper, at a fraction of the cost of discovering them after the exchanger has been bought.

What the duty is, and what it fixes

The duty is the heat-transfer rate the exchanger must deliver. It is set by what the process needs — cooling a stream from one temperature to another, recovering heat from warm waste water, condensing a vapor or boiling a liquid — and once the duty is fixed, everything downstream is fixed too: the stream flow rates, the outlet temperatures, the heat-transfer area required, and the pumping power needed to push the fluids through the unit.

A duty reaches the designer in two ways. Most often it is a process requirement computed from a specification on one stream; sometimes it is reconstructed from plant measurements. In either case there are always two independent routes to the same number, because the exchanger — ideally adiabatic, exchanging no heat with its surroundings — lets the heat the hot fluid gives up pass entirely to the cold fluid, less only a small loss to the environment. Two independent energy balances, one on each stream, must agree. When they do not, something in the data is wrong, and it is much cheaper to find out on paper than after the exchanger is bought.

The energy balance on each stream

For sensible heating or cooling — no phase change, approximately constant specific heat — each stream obeys the same equation:

Q̇ = ṁ · cp · ΔT

Here Q̇ is the heat-transfer rate in kilowatts, ṁ the mass flow rate in kilograms per second, cp the specific heat capacity — the energy needed to raise one kilogram of the fluid by one degree — and ΔT the temperature change of the stream. On the hot side the temperature falls, so the duty equals ṁ·cp times the inlet-minus-outlet difference; on the cold side the temperature rises, so it equals ṁ·cp times outlet minus inlet. The specific heat must be evaluated at the mean temperature of the stream, because real fluids change their cp with temperature; taking it from a table at the wrong condition is one of the classic ways a duty goes wrong.

Two ideas build directly on the balance:

• Heat capacity rate. The product C = ṁ·cp, in kilowatts per degree, is the stream’s heat capacity rate — the quantity that governs how much its temperature changes. Rearranged, the balance reads ΔT = Q̇/C: for the same duty, the stream with the larger heat capacity rate changes temperature less. Saying the two duties are equal is exactly saying that C_hot·ΔT_hot = C_cold·ΔT_cold.
• The temperature diagram. Plotting each stream’s temperature against the cumulative heat transferred gives two curves whose vertical separation is the local driving force. The diagram shows the whole duty at a glance — the size of each temperature change, where the streams run close together, and how the driving force is distributed along the exchanger.

A worked example: one duty, computed twice

The following is a hypothetical example. Consider a water-to-water cooler: 2 kg/s of hot water is to be cooled from 90 °C to 60 °C, using 3 kg/s of cooling water that enters at 20 °C. Take the specific heat of liquid water as roughly 4.18 kJ/(kg·°C), its approximate value at moderate temperatures.

Hot side. Q̇ = 2 kg/s × 4.18 kJ/(kg·°C) × (90 − 60) °C ≈ 251 kW.

Cold side. The cold stream’s heat capacity rate is 3 × 4.18 = 12.5 kW/°C. If the balances agree, the cold water must rise by 251 ÷ 12.5 = 20 °C and leave at about 40 °C. Computing the duty from measured data on either side should return the same figure of roughly 251 kW.

The table also shows the internal consistency the balance enforces: the hot stream has the smaller heat capacity rate, so it changes temperature more — 30 °C against 20 °C — yet the products C·ΔT are identical, as they must be.

The mismatched case is just as instructive. Suppose the cold outlet had instead been logged as 46 °C. The cold side would then claim 3 × 4.18 × 26 ≈ 326 kW of duty, which cannot be reconciled with the 251 kW the hot side surrenders. Both cannot be right, so something in the data is wrong — an inaccurate flow meter, a mis-calibrated temperature sensor, a specific heat taken at the wrong condition, or a heat loss nobody has accounted for.

Why measured duties disagree

In an ideal adiabatic exchanger at steady operation, heat lost by the hot stream equals heat gained by the cold stream exactly. Real measurements rarely balance perfectly, and the size and pattern of the gap point to its cause:

• Heat loss to the surroundings. Imperfect insulation leaks a little heat to the environment, so the cold side consistently reads slightly below the hot side.
• Property variation. Specific heats vary with temperature and composition; a cp taken at the wrong condition shifts one side’s duty systematically — a duty effectively measured with the wrong specific heat.
• Measurement uncertainty. Flowmeters and temperature sensors each contribute error, and the errors compound in the product ṁ·cp·ΔT.
• Unsteady operation. During startup, shutdown, or load changes, energy is being stored in the equipment and the fluid inventory, so instantaneous duties on the two sides need not match.
• Unrecognized phase change. A stream that is partly condensing or vaporizing has a duty dominated by latent heat; the sensible-heat formula then understates its duty drastically.

A disciplined recheck resolves most gaps: recompute cp at each stream’s mean temperature; confirm the flows and temperatures being compared were taken at the same time under steady conditions; verify meter calibration; and examine the pattern of the disagreement. A consistent bias in one direction suggests heat loss or a property error; random scatter suggests instrument noise; a gap of tens of percent suggests a phase change or a gross data error.

Phase change: latent heat replaces cp·ΔT

Condensers and evaporators need a different energy term. A condensing vapor gives up its latent heat at essentially constant temperature, so the sensible formula — built on a temperature change — does not describe that side. The duty is instead Q̇ = ṁ·λ, the mass rate of condensation or vaporization times the latent heat λ. For steam condensing at atmospheric pressure, λ is about 2257 kJ/kg, so a steam heater’s steam consumption is simply the process duty divided by the latent heat, with sensible terms added if the condensate is subcooled or the steam arrives superheated.

The two-balance check carries over unchanged. A heater condensing, for example, 0.11 kg/s of atmospheric steam delivers roughly 0.11 × 2257 ≈ 250 kW, and the process fluid on the other side must show the same duty when its own energy balance is computed; any shortfall reappears as an unexplained gap between the two sides.

After the balances agree: the road to sizing

An agreed duty is the first link in a design chain. Next, the convective film on each side of the wall and the conduction through the wall combine into an overall heat-transfer coefficient U. Because the temperature difference between the two fluids changes along the exchanger, a special average — the log mean temperature difference (LMTD) — links the duty, U, and the required area. Counterflow arrangements sustain a larger average driving force than parallel flow, which is one reason they are usually preferred, and the effectiveness-NTU method takes over when outlet temperatures are not all known in advance; allowances for fouling and the choice of flow arrangement complete the sizing picture. These methods deserve treatment of their own, but the duty and its two energy balances come first, because every number downstream depends on them.

Practical takeaways

1. Compute the duty independently from both streams with Q̇ = ṁ·cp·ΔT, evaluating cp at each stream’s mean temperature.
2. Expect the two duties to agree in steady operation: hot-side duty equals cold-side duty, less only a small loss to the surroundings.
3. Chase any mismatch in order — specific heat, concurrent flows, sensor calibration, steady state, hidden phase change, heat loss — before blaming the exchanger.
4. Use Q̇ = ṁ·λ for any stream that condenses or boils, and keep the balance check on the phase-change side too.
5. Proceed to the overall coefficient, the LMTD (or effectiveness-NTU method), and the area only once the duty is confirmed by both sides.