A take-or-pay clause obligates the off-taker (the site) to either consume, or pay for, a defined minimum amount of solar generation each year, regardless of actual on-site electricity demand. It exists to protect the developer's revenue from changes in your business that reduce consumption.
The mechanism is straightforward. If your annual generation is 1,000 MWh and the take-or-pay level is 90%, you must pay for at least 900 MWh of solar electricity each year. If your actual demand falls (say due to operational changes, site closure, reduced output) and you only consume 700 MWh, you'd still be billed for 900 MWh at the PPA rate. The remaining 200 MWh is exported to the grid and the developer keeps that revenue too.
Why this exists: developers and funders need underwritable cashflows. If you could simply stop consuming, the project economics fall apart. Take-or-pay shifts demand risk from the developer to the off-taker.
Typical levels: 85-100% of expected generation. 100% is the highest commitment and produces the best PPA rate; 85-90% is more flexible and slightly higher rate. Below 85% gets harder to underwrite.
The "% of generation" methodology means the obligation flexes with actual generation, if a cloudy year produces less, your obligation reduces proportionally. The "fixed kWh" methodology sets an absolute number you must take regardless of weather, which is harder to underwrite for the off-taker but possible.
Negotiation levers: shorter PPA term (15-20 years vs 25), stronger covenant or parental guarantee, willingness to accept a slightly higher start rate, or a step-down on take-or-pay over time.