The safety margin sets a buffer below your Contracted capacity limit (GTV) so that Tibo's control system — Alice — has room to respond before your site risks a GTV breach.
What the safety margin does
Your GTV is the maximum power your site is contractually or technically allowed to import from the grid. The safety margin reserves a portion of that capacity as a buffer, so Alice targets a lower effective limit during normal operation.
Example: If your GTV is 160 kW and you set a safety margin of 20 kW, Alice will manage your site to a target of 140 kW — leaving 20 kW of headroom before a breach would occur.
A smaller margin means Alice can use more of your available grid capacity, but it also reduces the window available to respond to sudden changes in load or generation. A larger margin is more conservative but sacrifices some usable capacity.
Factors that constrain how tightly you can set the margin
The right margin is not a fixed number — it depends on the characteristics of your site. Three main factors determine how much headroom you realistically need.
Phase imbalance
Most grid connections are three-phase, and power is rarely distributed evenly across all three phases. If your site has significant imbalance — where one phase consistently carries more load than the others — the GTV can be breached on that phase even when your total site power looks acceptable.
A tight margin based on total power gives a false sense of safety if one phase is disproportionately loaded.
Sites with high phase imbalance need a margin that accounts for the worst-case phase, not just the aggregate.
This is relevant in both simulation and Live EMS contexts — check per-phase data when reviewing your load profile before deciding on a margin.
Asset response speed
Alice can only control a GTV breach as quickly as your assets can respond to its commands. Slow-responding assets — batteries with long ramp times, for example — cannot react instantaneously when load rises sharply.
If the margin is very tight, there may not be enough time for a slow asset to ramp up or curtail before the GTV is breached.
Faster-responding assets (such as certain inverters or direct load controls) can support a tighter margin more reliably.
In a simulation context, asset response speed may be modelled as ideal — results in simulation may therefore look better than real-world performance. Apply additional caution when translating a simulated margin to a live configuration.
Unpredictable load spikes
Some sites have consumption patterns that are difficult to predict — industrial equipment switching on suddenly, EV chargers starting simultaneously, or other intermittent high-draw processes.
Sudden load spikes can push the site toward the GTV faster than Alice can respond, regardless of asset speed.
Sites with irregular or spiky consumption profiles need a larger buffer to absorb those events safely.
In Live EMS, you can observe historical spike behaviour in your load data to inform the margin. In simulation, spike behaviour depends on the quality and representativeness of the input data — be cautious if the simulation dataset is smoothed or averaged.
The margin is site-specific
Because these three factors vary significantly between sites, there is no universal correct margin. The right value requires an understanding of:
Your per-phase load distribution
The response characteristics of the assets Alice is controlling at your site
Your typical and worst-case load profile, including any known spike events
Setting the margin without this context — even if a lower number looks attractive on paper — carries a genuine risk of GTV breaches in operation.
Before reducing an existing margin
If a margin was set during an earlier Tibo configuration, it is worth investigating why that value was chosen before reducing it.
Reducing the margin without understanding the original rationale increases the risk of breaches, particularly during edge-case operating conditions that may not appear often in historical data.
Where possible, review at least several months of load data across all phases and under varied operating conditions before committing to a tighter margin in a live environment.