An automated system does not start or stop gently: it draws power in peaks, and what decides whether it runs well is less how much power you have contracted than how stable the energy reaching the switchboard actually is. Before automating, STOKA surveys the warehouse electrical system alongside every other building condition. Here is what gets measured, who measures it, at what point in the project, and what alternatives appear when the available power falls short. There is no minimum that rules your warehouse out: there is a design that adapts to what you already have.
Why the electrical survey comes first, not last
Power is one of the few building conditions that can move an entire project schedule. A supply upgrade is negotiated with the utility and runs on its own timetable, which depends neither on STOKA nor on the client. If that process starts once the equipment is already in the country, the sequence breaks down: installing an automated system takes 3 to 4 months from the purchase order (STOKA figure), and that window assumes the electrical work advances in parallel, not afterwards. That is why the electrical survey happens during the diagnostic stage, together with clear height measurement, floor condition and the real flows of the operation. It is not a closing formality: it is one of the inputs to the design. If the supply constrains something, you want to know while the system can still be sized differently.
How much electrical power does an automated warehouse need?
This is the most searched question on the topic and it has no single answer. Consumption depends on the equipment you end up installing: racking served by stacker cranes, the machines that travel an aisle and raise and lower the load, does not draw the same as a fleet of shuttle robots, and a pallet warehouse does not draw the same as a carton one. The number comes from the datasheet of the chosen equipment and from the real duty cycle: hours per day, movements, starts per hour. Publishing a generic figure would mean inventing it, and nothing here is estimated. What can be defined before choosing equipment is what has to be measured in your warehouse so the calculation is later made with data rather than assumptions. That is the entire purpose of the electrical survey.
What gets checked at the switchboard and the service entry
The survey records the basics first: contracted power and actually available power, which are almost never the same number, because the warehouse already has its own loads such as lighting, doors, compressors and forklift battery charging. Next comes the type of service entry and whether supply is single phase or three phase. At the switchboard, the team reviews spare capacity for new circuits, the condition and selectivity of the protective devices, and whether there is physical room for new boards near where the system will sit. Earthing is measured, never assumed: resistance value and continuity along the whole path are verified, because both personnel safety and control electronics depend on it. The survey also records where cable routes can run without crossing traffic aisles or the loading yard.
Power quality: voltage dips and micro outages
This is the point that does the most damage and gets the least attention. Automated equipment copes reasonably well with working near the power limit; what it does not cope with is dirty energy. Voltage dips on starting, micro outages lasting fractions of a second, phase imbalance and harmonics injected by variable speed drives all translate into controller restarts, lost position references and stoppages that later get logged as equipment faults when they are in fact installation faults. That is why verification cannot be a snapshot: behaviour is logged over several days of normal operation, with the plant working and through the heaviest shifts. A one hour reading on a Saturday morning says nothing useful. That log is also the input that later determines what needs correcting and what does not.
Backup: what happens if power fails with a pallet mid travel
In a manual warehouse, an outage stops the operation. In an automated one, an outage can leave a load unit halfway into a rack that reaches considerable height, and getting it out of there takes time. Backup is therefore sized to finish what is in progress and shut down in an orderly way, not to keep producing. The survey defines what needs uninterruptible supply, usually a UPS or uninterruptible power supply feeding controllers, the servers of the warehouse management system (WMS, the software that decides what is stored where and in what order each order is picked), the network and the security cameras; and what can wait without consequences. If the warehouse already has a generator set, its capacity and transfer time are verified. If it does not, the need for one is assessed.
What happens if the available power is not enough
Payback is calculated using the full investment and your operation’s net savings. Separate genuine expense reductions from capacity available for growth; freeing space or time does not automatically save cash.
When the electrical system is not the problem
It is worth saying plainly: there are warehouses where the installation is faultless and automating still does not make sense. If the real problem is that stock sits in the wrong locations, that there is no cycle counting, that orders are picked without a sequence, or that the layout forces unnecessary travel, no amount of power fixes it. In those cases the right order is process first: reorganise locations, set rotation rules, measure movements properly, and only then consider whether an automated system adds anything. Surveying the electrical system anyway pays off twice: the data is ready for when the moment comes, and several of the energy improvements that surface along the way pay for themselves without automating anything. It is also what later makes heights of 7 to 40 metres (STOKA criterion) possible without surprises.
In short: energy gets verified, not assumed
The electrical installation does not decide whether your warehouse qualifies for automation: it decides how the system is sized and in what order the stages happen. What gets measured is contracted and available power, the switchboard, the service entry, earthing and power quality across several real days of operation. The person who verifies and signs it off is a licensed electrician or the electrical designer, not the equipment supplier. And if the power falls short, there are ways out: upgrade the supply, free up load, or design the system differently. The only thing that never helps is finding out late.
Frequently asked questions
A licensed electrician or the plant electrical designer, carrying professional responsibility for what they sign. The equipment supplier does not verify it: the supplier provides the datasheet with each machine consumption and requirements, and the licensed professional compares that against what is installed and defines what has to be adapted.
Supply type is one of the first items surveyed, because it shapes the sizing and may require arrangements with the utility. It is not a filter that rules the warehouse out: it is a design variable. Depending on the chosen equipment and the duty cycle, the supply may need converting, upgrading, or the automation scope may be narrowed instead.
It depends on the utility, the area and the size of the upgrade, so there is no single lead time worth publishing. What matters is sequence: since installing the system takes 3 to 4 months from the purchase order, the electrical process has to start in parallel or earlier, never once the equipment is about to arrive.
Not always. Backup is sized to finish movements in progress and shut down in an orderly way, not to keep operating through the outage. Uninterruptible supply on controllers, servers and the network is often enough. A generator set is assessed when the operation cannot stop, or when the area suffers frequent and long outages.
It is an interruption or voltage sag lasting fractions of a second, so brief that lights barely flicker and nobody notices. For an automated controller it is long enough to restart and lose its position reference. That is why power quality is logged across several days of normal operation rather than measured once.