02
Can we harness the storm?
A question I carried for roughly twenty years is finally being treated like an engineering problem rather than a wish: write down the assumptions, model the physics, let the results break the idea, and use each failure to decide what deserves the next experiment.
The first simulations asked whether part of a lightning transient could be captured while an independent path still protected the system. Those early models taught us quickly that “captured” and “retained” are not the same thing. Energy initially stored in the capacitor leaked back through the permanent ground path, so the next experiment introduced one-way isolation. That solved the retention problem in the ideal model, but immediately raised harder questions about forward drop, leakage, switching time and the unrealistic ideal lightning source.
Successive versions added those imperfections, replaced the ideal source with a finite-impedance source, introduced controlled overvoltage diversion, measured component stresses and tested layered protection. Some ideas survived. Others did not. Layered protection, for example, added complexity without producing a clean overall advantage over the best comparable single-stage case.
The largest wall arrived in v0.2.0, when the simulation requirements were compared against documented physical protection technologies. After primary-source verification corrected several pieces of component evidence, the central result still held: the shared protection-and-harvesting terminal was rejected as specified. The strongest conflict was not simply that lightning is “too powerful.” Protection wants to divert the transient aggressively, while useful harvesting wants enough coupling to transfer energy. Asking the same electrical node to perform both jobs created incompatible requirements in the architecture we tested.
That failure changed the question.
At that stage, the research stopped assuming that the entire strike should be captured, or that every useful joule must remain electrical. The next proposed ideal study asked whether a deliberately limited fraction—initially 25% as an experimental target, rather than an achieved result—could be admitted to a harvesting system while protection retained priority. The admitted portion could then be divided between useful electrical recovery and deliberately captured thermal energy. The requested fraction would eventually be swept upward and downward to find a feasible and worthwhile frontier rather than maximizing capture blindly.
That led to another idea: perhaps the missing interface is not one giant storage component at all. Computer networks buffer bursts. Hydraulic systems provide pressure transients somewhere else to go. Vascular systems and river deltas distribute concentrated flow across many paths. Shock absorbers transform an impulse over time. These analogies do not prove an electrical solution exists, but they suggest a testable abstraction: take energy concentrated in space and time and distribute it across space, time and, potentially, useful energy forms.
Pulsed-power engineering makes that analogy worth investigating seriously. Real systems already use distributed capacitors, transmission lines, intermediate energy stores, impedance transformation, parallel modules and precisely timed switching to shape enormous electrical pulses. The open question is whether any physically admissible version of those principles can work in the opposite problem: accepting an uncontrolled lightning transient, admitting only a protection-compatible portion, and spreading that energy without recreating the same protection-versus-harvesting conflict.
The project remains deliberately falsifiable. A failed architecture stays in the record. A weakened argument gets corrected. A promising analogy still has to survive equations, simulation and physical evidence.
The objective is not to prove that lightning harvesting works. It is to discover what would have to be true for useful storm-energy recovery to work—and whether nature and engineering actually permit it.
RESEARCH UPDATE · OCTOBER 21, 2026
Where can diverted energy become useful?
The project keeps two connected questions open: can some electrical energy be captured and retained, and can energy being diverted be harnessed in another useful form? The current focus remains on the second question: investigate diversion arrangements that preserve protection, then determine whether useful energy can be recovered along that route.
Heat is one candidate. Electrical recovery, magnetic coupling or storage, and mechanical conversion are also questions to investigate. None is assumed feasible merely because the energy exists. Each candidate must show where the energy goes, what can actually be recovered, and how recovery affects the diversion path.
The resistor bank was one tested approach, not a required component of every future design. Other conductor arrangements and energy-transfer mechanisms remain open for investigation. A conduit or conductor route is a design question; a route alone does not establish a useful recovery mechanism. The next step is to explore candidate mechanisms before choosing the next bounded experiment. Once this diverted-energy investigation reaches a satisfactory, evidence-supported conclusion, we will return to exploring direct electrical capture and retention.
One branch explored nested conductive shells as a capacitor. The source-connected simulations gave that idea a useful reality check: the shell candidates stored only tens of millijoules at the modeled operating voltage and added essentially no benefit to the baseline resistor-bank response. That result belongs in the record. A striking geometry still has to earn its place in the system.
Checkpoint 3 compared three approaches: sending admitted energy into a resistor bank, adding a passive buffer, and modeling controlled admission with a slower release into the bank. The controlled-release study showed why spreading a pulse over time is worth investigating. In a representative comparison, peak bank power fell from about 24.8 megawatts to 100 kilowatts, while total bank heat fell from about 3.20 to 2.68 kilojoules. The switch and converter remain ideal requirements models; suitable real hardware has not been demonstrated.
That distinction matters to the direction of the project. A lower peak can make the delivery problem more manageable, but it does not create additional energy. We also have to distinguish heat generated in the resistors from heat collected, retained and delivered to a useful process.
The thermal model brought another constraint into focus. In the representative baseline, about 2.68 kilojoules reached a five-kilogram water store, whose peak temperature rose by roughly 0.14°C. None of the tested cases reached the assumed 40°C useful-load threshold. These are simulation results for the tested configurations, rather than a general limit on all storm-energy recovery. They do show why a dramatic electrical pulse does not automatically become a useful hot-water supply.
A question for the thermal path: can recovered heat produce collected water?
A specified follow-up for the thermal path follows that heat into a seawater evaporation and condensation model. The larger idea is a coastal seawater reservoir feeding a smaller process stream. That proposed thermal study is bounded: resistor bank, isolated heat transfer or storage, a small evaporator, a condenser and a collection tank, with concentrated brine accounted for separately. It remains one possible follow-up within the broader diverted-energy investigation.
We will track the temperatures, equipment heat capacity, transfer losses, vapor formed and liquid actually collected. Having enough joules on paper is only part of the question; the heat must also be available at a temperature that supports the process. The existing bank remains below atmospheric boiling temperature after a single modeled stroke. A smaller quantity of water alone cannot remove that temperature constraint.
A result of no boiling or no collected condensate is a valid outcome. Any redesign will have to show what changes in temperature, thermal mass, energy capture or accumulation are required, while preserving the electrical constraints. Repeated-event scenarios must include cooling between events and recovery of the electrical system.
Direct electrical capture remains its own research path. Separately, a later thermal-path question is whether some recovered heat could be converted back into electricity. That conversion would require a suitable temperature difference and would share the thermal budget with any water-production process. Energy retained electrically, dissipated, recovered as heat and exported as work must all be accounted for once. Any modeled condensate would also need later water-quality assessment before it could be described as drinking water.
This is where the project is heading: explore how electrical energy can be diverted while preserving protection, and whether useful energy can be recovered in electrical, thermal, magnetic or mechanical form. The current resistor results help evaluate one approach; they leave other mechanisms open. Candidate ideas must earn their results through energy accounting, component constraints and reproducible tests. The work remains computational research. The evaporation-and-condensation experiment is specified as a thermal follow-up; this update reports no completed desalination result or physical installation.
Current focus · diverted-energy recovery · electrical, thermal, magnetic and mechanical possibilities under investigation