KNOWLEDGE / SOLAR + STORAGE The short answer AC coupling connects solar and storage through separate converters on an AC bus. DC coupling connects them before a shared inverter. The best choice depends on charging paths, equipment constraints and the operating job. Hero: CurrentCells original explanatory diagram. Simplified power paths, not a construction drawing or a photograph. A solar farm and a battery can occupy the same land, use the same grid connection and still be electrically very different projects. The crucial question is where their power paths meet: on the alternating-current side of the equipment, or on a direct-current connection before electricity passes through a shared inverter. That choice is called AC or DC coupling. It shapes which energy the battery can capture, which equipment the plant shares, and whether solar production and battery discharge compete for the same conversion capacity. It does not describe the battery chemistry, guarantee backup power, or tell an investor which project will earn more money. The useful comparison starts with a wiring diagram and an operating objective, not a claim that one architecture is always better. For a new solar-plus-storage plant, capturing otherwise clipped energy may matter. For an existing solar installation, preserving working equipment may matter more. For both, the grid connection can remain the binding limit. What actually gets coupled Solar photovoltaic modules generate direct current. A battery also accepts and delivers direct current at its terminals. The public electricity grid operates on alternating current, so power electronics must connect those DC devices to the AC network. An inverter converts DC to AC; a bidirectional battery inverter can also convert AC to DC during charging. In a typical AC-coupled system, the solar array has its own inverter and the battery has a separate bidirectional power conversion system. Their outputs meet on an AC bus. To store solar production, electricity travels through the solar inverter, across the AC connection, and through the battery converter before reaching the cells. In a typical DC-coupled design, the solar array and battery meet on the DC side of a shared bidirectional inverter. A DC-to-DC converter commonly manages the battery connection and voltage relationship. Solar energy can reach the battery without first making a trip onto the AC bus. Later, battery discharge goes through the shared inverter to supply the grid. These are simplified topologies, not construction drawings. Actual products differ in conversion stages, voltage ranges and controls. Co-location only says the assets are near each other. It does not establish whether they share an inverter, a controller, an interconnection agreement or even an owner. Why DC coupling can capture clipped solar A solar array is often rated for more DC power than its inverter can export as AC power. The larger array can improve inverter use during weaker sunlight. At some sunny moments, however, potential DC production exceeds what the inverter can process. Without another useful path, the plant limits production. That is inverter clipping. A DC-connected battery can sometimes absorb that otherwise unavailable energy before it reaches the constrained inverter. This is different from asking an AC-connected battery to recover electricity after clipping has already occurred. Once the solar inverter has limited the DC power it takes from the array, that forgone production is not sitting on the AC bus waiting to be stored. The word sometimes matters. The battery needs spare charging power, usable capacity and an allowed operating state. The converter and DC connections must support the transfer. A full battery cannot keep absorbing midday surplus, and a battery reserved for another service may not be available to do so. Clipping is also not a synonym for every kind of curtailment. An export restriction at the grid connection may leave AC solar energy available to charge an AC-coupled battery behind that restriction. A command that shuts down generation, a voltage constraint and a DC inverter bottleneck are different conditions. Ask where the lost energy occurs before attaching a recovery claim to it. A worked example: one export limit, different power paths Consider a hypothetical plant with 120 MW of available solar DC power at one instant, a 100 MW shared inverter and a battery capable of charging at 30 MW. To make the power paths clear, ignore conversion losses, auxiliary consumption and voltage limits for this example. These are illustrative numbers, not operating data from a real project. With DC coupling and enough battery headroom, the plant could send 100 MW toward the grid and 20 MW into the battery. Without a storage path on the DC side, the same 100 MW solar inverter could not convert all 120 MW at that instant. The 20 MW difference illustrates potential clipping recovery, not a guaranteed annual energy gain. Now consider a later hour with 80 MW of solar available and a request for 40 MW of battery discharge. Through the same 100 MW shared inverter, the combined 120 MW cannot all reach the AC side. The controller must reduce one contribution or otherwise manage the constraint. The battery rating does not automatically add to the solar inverter rating. Separate AC inverters could remove that particular shared-inverter bottleneck, provided both were adequately rated. But if the whole site has a 100 MW export limit, the combined export would still be limited to 100 MW. Separate converters do not create extra interconnection rights. Finally, power is not energy. Charging at 20 MW for half an hour would put 10 MWh into the idealized storage path before losses. Whether that opportunity occurs for a few minutes or hundreds of hours each year determines its significance. A single sunny-hour diagram cannot establish the value of a storage investment. Fewer conversions do not mean a universal efficiency number Avoiding an AC conversion on the solar-to-battery path can reduce losses. The 2024 Annual Technology Baseline illustrates this with assumed round-trip efficiencies of 87 percent for charging its representative DC-coupled battery from solar and 85 percent for charging from the grid. Those figures are modeling assumptions for that defined system, not a universal specification or a current equipment warranty. The distinction is important because efficiency depends on the route being measured. Solar-to-battery-to-grid energy follows a different path from grid-to-battery-to-grid energy. A cell efficiency figure excludes equipment that a project-level meter may include. Cooling, standby demand, transformers and converter loading can all influence delivered energy. Compare systems using the same input and output measurement boundaries, charging source and duty cycle. A supplier quoting DC-terminal efficiency is not necessarily offering a more efficient power plant than one quoting net AC delivery at the grid meter. CurrentCells has a separate round-trip efficiency guide for that measurement problem. Clipping capture also needs its own estimate. The same 2024 baseline found only 0.0 to 0.5 percent of DC production lost to clipping in most of the contiguous United States for its representative roughly 1.3 inverter loading ratio. It therefore treated clipping recovery as negligible for that particular configuration. Higher loading ratios can change the opportunity. The lesson is to simulate the actual array and site, not borrow a generic recovery percentage. Why retrofits often favor a different answer Adding storage beside an existing solar plant is not the same decision as designing both assets together. An AC-connected battery can often join the existing AC infrastructure without replacing the solar inverter architecture. That can make it attractive where preserving equipment, warranties and operating arrangements is valuable. It is not automatically a plug-in upgrade. The transformer, s