New backup power sensor or timer: pumping tank to tank

[zerohourrct]

I was brainstorming better/alternative ways to signal backup generators to turn on using pumps. I've researched three previous designs and I'd like to introduce a fourth design, which also has applications as a timer, all due to the mechanics of pumping in a closed loop from tank to tank.

My new Approach 4: Tank to tank: uses two tanks and a single pump between to create a difference in quantity between tanks equal to maximum pump head (wikipedia). This maximum pump head is directly proportional to pump power: if the pump turns off it will be zero, if an overload condition occurs the difference in capacity will shrink directly proportional to resulting pump power (this slightly contrary to real life phisycs where ideal pump head delta would be the squared cube root of applied power delta). This output signal could be calculated by two arithmetic combinators, or simply directly output to hot-water pumps that will turn on the backup generator at a preset value, I recommend about 2-5 units different from whatever equilibrium value a 100% powered pump creates. The delay time in returning to equilibrium max pump head gets exponentionally or logarithmically longer the closer to equilibrium level the tanks get. This can be combined with a reset latch to produce a time delay circuit. Adding additional parallel tanks to the 'high' side will further lengthen this recovery time. The pump head seems proportional to the length of pipes in the tank return path: the smallest path of 4 pipes by using pipe-to-ground results in 140 units of max pump head. A standard straight length of 7 pipes gives 150 units max pump head. See below pictures using a pipe length of 7.

tanktankhigh.PNG (305.47 KiB) Viewed 5130 times

tanktanklow.PNG (251.09 KiB) Viewed 5130 times

Approach 1: Smart chest, inserter, belt sensor: uses a smart chest and two inserters, one is powered from a constant reference source, the other is powered from sensed source. When power is 100% the two inserters maintain a constant quantity inside the chest. As the sensed source power experiences overload throttling or a complete loss of power, the sensor inserter slows down or stops, resulting in a change of quantity inside the chest which is the output signal. Advantages: very simple and space efficient. Disadvantages: sometimes unpredictable, requires some manual effort to setup or fix if it breaks, requires a constant power source, very coarse signal output, hard to set intermediate trigger levels.

Approach 2: Cold-water pumps: uses two tank and two small pumps in parallel with a standard 1:14:10 generator setup to provide a back-flow of cold water through 10 steam engines. A similar method was used as a bootstrap opamp by MadZuri. The coldwater method exploits two mechanics. First, that pumps pumping straight out of a tank has flow priority in a fluid network: as long as the pumps have full power they will pump the full flow of cold water and zero hot water can flow through the parallel boilers. Second, steam generators will happily consume 6 units of cold water and produce zero output, as long as there is power demand for their operation. As soon as a power loss OR overload throttling occurs, the pumps lose pumping power, and hot water is shunted through the alternate boiler path to power the engines. Advantages: very simple, no manual setup required. Disadvantages: overload throttling is somewhat unpredictable and will not make up to 100% conditions, no actual sensor output of power condition, hot water can become shunted back into the water tanks if the steam turbines stop eating 6 units of water due to electric demand, unless 2 additional check valve pumps are used.

Approach 3: Xknight's power indicator sensor: complicated sensor design that abuses power priority, several pumps, and dummy loads, to create a tank level proportional to accumulator level. Advantages: highly precise signal output of accumulator level. Disadvantages: False output if new accumulators are being added to network, requires significant manual setup to get water levels correct, complicated, requires dummy loads, does not sense overload conditions.

[zerohourrct]

What I'm researching now is a sensing circuit that determines if the rate of accumulator discharge is too high to last through the night or exceeds the current capacity of solar panels, and turn on backup generators to make up the difference. I could compare the current accumulator capacity with a constant 'worst-case' accumulator level, or I could directly compare the rate of discharge with the worst case rate.

[provet]

Interesting read, thanks! will def. take a look into this

[hansinator]

Nice setup! I needed a way to determine how much power there is to shut down my smart furnace safely before a blackout. This helps a lot. The only downside is the manual setup needed. I wish there was a solution that just works when you place a blueprint.