Photovoltaic panels produce DC current that heats a typical water heater

Simply put, the goal is to heat a standard hot water tank with solar PV. To be able to help me you'll need to understand (among other things): DC power and loads, heater resistances and solar PV optimization (resistance as it relates to fluctuating PV voltage), and Tigo DC optimizers. Please note this project v1.0 is already working and has been working for over a year, I just know it can work better. This next phase involves increasing the performance as well as increasing the safety/control aspect of it (ie: it currently has no means to automagically shut itself off). The 6-panel array produces 1,260W @STC (see attached PV spec sheet for Vmp Imp and be sure to refer to specs for the 210W model). Each panel in the array has Tigo Optimizers attached to it. Under normal conditions on an average day I'm seeing power production anywhere from 500W to 1kW with sun, depending on the season and the weather conditions of course. Each PV panel plugs into it's own Tigo optimizer (so there are 6, one for each panel), and then all those optimizers are wired in series and that series terminates in a weather tight box (2 wires, one positive and one negative) where it connects to buried TECK90 #10 AWG cable (conductors are stranded copper, TECK is rated for 600V). This underground cable runs approx 150 ft back to the house. In the house it terminates inside a large metal box (nice and big, was once used to hold a simple 6"x6" PCB MOSFET setup which I can tell you more about) where it connects to #14 AWG (solid copper conductors). That #14 travels less than 8 ft with the two wires terminating onto the bottom heating element of a standard run-of-the-mill north american hot water heater. The water heater thermostat has been bypassed since they only work with AC, and this is DC. This tank is a pre-heater and is one of two heating tanks in the house, the second and final tank is a typical gas heated tank. There are 4 plumbing valves that allow me to isolate one of the tanks for the house hot water supply. The Tigo aspect of the system is of great assistance because not only does it maximize each PV panel, it also provides me with technical data and analysis tools via their monitoring portal website. I can give you access to this no problem, you will see data (power, volts, current) for both live data as well as the past year of graphed data. My goal is to study the performance shortcomings or "blips" that I see on the Tigo portal. Secondly, to increase the safety/controlling aspect of it. The "blips" are either 1) not real, or 2) are how the Tigo's respond when connected directly to a DC load. I notice on the Tigo portal that although it is working, random panels seems to fall out and lose 20% power and I wonder if I'm missing something in my design (a resistor, a capacitor?) to smooth things out. Anyways, diagnosis is required. And less important for this phase of the project, the safety aspect -- the water heater has a pressure release valve on it so technically there is some protection if it boils, but I'd like you to suggest some kind of component that's not too expensive to help with this. Basically to shut things down if the tank gets hot enough. Tigo already has a way for the MMU to communicate with the optimizers to tell them to shut themselves off, but at the moment it is human initiated as the only way to do it is to press the red "PV-SAFE OFF" button on the MMU. Any suggestions on this are appreciated. Or maybe a DC relay is the way to go, but they're expensive. In conclusion it is working and on good days the 60 gallon tank will go from cold to over 100 deg F, so the proof of concept is done and it works. I'm asking for your help to make it better. Your experience and knowledge of electrical theory and components will improve this and help me get to the next level. No tinfoil hats allowed LOL.