The effect somehow looks to be of the "avalanche" type - so after exceeding certain temperature - a steady increase in released energy is obvious.
Interesting that he explains the test-run of his 1MW plant limited to just 500kW because of "self sustaining mode".
It might be quite difficult to control that effect.
A serious point might be the thermal capacity of the setup.
By having a steady flow of medium he can increase that thermal capacity.
If the reactor starts to exceed the rated performance he can cool it down by increasing the flow of the medium.
As long as he is controlling the effect by means of resistive pre-heating with reasonable but lower cop - it might be easier to control the effect.
If he wants to go self-sustained - the only way to control it would be the amount of energy transfered by the medium.
So this would mean that he would need more flow control and more efficient heat exchangers to control such device self-sustained.
What do you think ?
In my view the thermal effect is definitely an avalanche type.
This means that to control the process and avoid melting of the Ni powder a few options are to be considered:
1)
Selfsustaining: the thermal capacity of a complete system is such that during too high heat production the cooling of the core is increased by increasing the (re)circulation of the cooling water, e.g. by increasing the pump rpm. This should be done such that the temperature of the core does not decrease too much as this will shut down the heat production process. Disadvantage of this option is that in case of a power failure, the core will melt due to a lack of cooling.
2)
Heat biasing: A permanent electrical heating is present (by an electrical powered heating element) during heat generation of the actual LENR process. As soon as the temperature of the core get too hot the bias heating is reduced. If temperature drops too low, bias heating is increased (also used for starting up the process). This allows simple control and easy shut down of the system: simply shut down the heat biasing. It seems that the Hyperion of Defkalion uses a pulsed modulated biasing method.
When designed with adequate dimensions, this method allows a LENR process that can prevent a core melt down in case of a power failure.
In my view, for safety and robustness reasons, Rossi uses option 2).
Disadvantage is that there is a permanent electrical power required to keep the ecats running and this lowers the COP.
Rossi indicates that the home ecats have a COP>6, hinting at using permanent electrical energy to control the ecats.