I will take a stab at an answer in what I hope might be simple and general terms.
What is defined as heat rate is akin to a relative inefficiency or relative loss rather than heat rate. The value itself has no units. Multiple by 100 and you get a percentage.
You first statement says, the greatest relative loss we expect in a plant is 10% over its minimum. The minimum relative loss occurs when the plant is operating at it greatest efficiency. So, a plant that is 85% efficient in the best case has a heat rate (relative loss) of $1/0.85 = 1.2$ or 120% in the best case. We expect this plant to have a heat rate (relative loss) of $1.1 * 1.2 = 1.3$ or 130% at its worst case operation.
Your next statements say, the worst case (130% heat rate) happens when the plant operates at its lowest capacity. Capacity is akin to the actual amount of electricity that is produced. So, a plant will be least efficient (have the highest heat rate) when it produces the lowest amount of actual electrical energy.
Your next statement says, the lowest capacity of a plant is 40% of its maximum capacity. Basically, this says, we do not operate power plants to produce less than 40% of their maximum rated output of electricity. Let's suppose the maximum capacity of the example power plant from above is $C_{max} = 100$ (with whatever units). The lowest capacity is $C_{min} = 0.4C_{max} = 40$. At 40, the plant operates with a heat rate of 130%. Now for the tricky part ... We apparently have no information about the relationship between increasing capacity and decreasing heat rate. We do not know whether the minimum heat rate of 120% is at maximum capacity ($C = 100$) or perhaps even before. All we know is, as we increase capacity, we can expect that heat rate decreases.
The finding leaves us with this conclusion. Absent any further information, we are free to assume any model we want for a relationship between increasing capacity and decreasing heat rate.
In real life, power plants are not operated continuously at their physical maximum capacity. They are operated below this physical maximum limit, by example continuing from above perhaps at $C_{operation} = 80 - 90$. In real life, inefficiencies also kick in above this operational design "set point".
My instinct to tackle this absent any other information would be to do the following:
- Set the maximum heat rate at the lowest capacity. This is defined by reference.
- Set the minimum heat rate somewhere at 80% - 90% of maximum capacity. This can be defined by further research (i.e. look up typical operating capacities of power plants relative to their rated maximum capacity).
- Define the relationship between increasing capacity and decreasing heat rate as linear over the above range. This is your model. Publish it.
- Do not model the behavior of systems below their minimum capacity.
- Do not model the behavior of systems above the 80% - 90% their rated maximum capacity.