A reading looks impossibly high (data spikes)

A spike is a single bad record in the stored data. Augos can remove it and recalculate the affected period. Tell support@augos.io the point and the date and they will investigate.

What you will see

• One day dominating the whole month, with the other days looking like zero. They are usually not zero, just tiny by comparison because of the chart scale.
• A maximum demand figure the site physically cannot draw.
• An estimated demand-based charge far above what the utility actually billed.
• A water meter jumping by two orders of magnitude while its neighbours look normal.

Why it matters more than it looks

Demand-based charges are set by a single highest half hour in the month. One spike anywhere in the record therefore sets the month's maximum demand, and that propagates into every kilovolt-ampere based charge in the estimate. A spike from months ago can still be distorting a cost estimate today.

What to do

1. Note the exact date and the point.
2. Check whether the surrounding days really are zero or just small. Chart scaling makes normal days look flat next to a spike.
3. Email support@augos.io with the point path and the date. Say why the figure is impossible for your site: installed load, shift pattern, what was running.

What Augos will do

Support reviews the historical data, identifies the spike, has it removed, and recalculates the affected period. Where the spike had already fed into a bill verification, that verification is recalculated too. Removing the spike alone does not refresh a verification that was already computed, so ask for both.

Good to know

• If one meter in a set jumps by orders of magnitude while its siblings look normal, treat it as a spike before treating it as a leak or a fault.
• Turnaround is usually same day for the removal. Where the spike has to go to the development team, expect a couple of days.
• A genuine consumption change looks different: it persists, and it usually has an operational explanation. See Our usage has jumped below in Good to know.

When it is not a spike

Before assuming bad data, widen the window. Three months is not enough to tell a fault from seasonality. A site whose consumption climbs every summer and falls every winter is showing you its heating and cooling load, not a metering problem. Ask support for a multi-year view of the point and the pattern usually answers itself.

Still stuck

Email support@augos.io with the point path, the date, and what you believe the figure should be.