There are exactly two ways to say how much heat an exchanger moves, and a rating is honest only when both agree.

The stream road: $\dot{Q} = \dot{m} \, c_p \, \Delta T$, read Q-dot equals m-dot c-p delta-T. $\dot{m}$ (m-dot) is the mass flow in kg/s, $c_p$ the specific heat in $\mathrm{kJ/(kg \cdot K)}$ — take 4.2 for water — and $\Delta T$ that stream's own temperature change across the unit. This road knows nothing about the metal; it is pure bookkeeping on one fluid.

The surface road: $\dot{Q} = U A F \, \Delta T_{lm}$. $U$ is the overall coefficient, $A$ the heat transfer area in $\mathrm{m^2}$, $\Delta T_{lm}$ the counterflow log mean, and $F$ is the LMTD correction factor: a bare number, never above 1, that derates true counterflow for a real shell-and-tube or crossflow arrangement in which some of the fluid inevitably flows the wrong way. $F = 1.0$ is true counterflow; below about 0.75 the arrangement is considered thermally unsound and the design is changed rather than fudged.

Sizing runs the two roads in order: the streams tell you the duty, and the surface equation, rearranged to $A = \dfrac{\dot{Q}}{U F \, \Delta T_{lm}}$, tells you how much metal that duty costs. Watch the one unit trap in the whole lesson: duties are quoted in kilowatts and $U$ in watts, and a thousandfold slip here produces an exchanger the size of a postcard.