Meana Petro

Digital & data

Remote site bandwidth

Tags, report rate and protocol turned into the kilobits per second a link has to carry — including the backfill surge after the link comes back.

Inputs

Video

512 kbit/s for a low-rate surveillance stream.

Outage recovery

In hours.

In hours.

Results

Loading the calculation…

This calculation ran in your browser. No value you entered was sent anywhere, and nothing was stored.

How it is calculated

  • Bytes per reported value by transport: about 14 for Sparkplug B, 42 for OPC UA binary, 90 for JSON over MQTT, 8 for a raw Modbus poll.
  • Report-by-exception suppresses values before they reach the wire, scaling the rate down linearly.
  • Backfill bandwidth is the backlog accumulated during the outage divided by the window you allow it to catch up in — this is the number that sizes the link, not the steady state.
  • The recommended plan rounds the peak up to the next 64 kbit/s increment, which is how satellite plans are actually sold.

Standards it speaks

Where this stops being true

This sizes throughput, and throughput is rarely what breaks a remote site. Geostationary satellite latency is 500 to 700 ms round trip regardless of how much bandwidth you buy, so anything that needs a control loop or an interactive session has to stay at the edge. Rain fade on Ku band takes the link out entirely for minutes at a time, which is a buffer problem rather than a bandwidth one, and contended plans deliver their committed rate and not their advertised one. Size the edge buffer alongside this number, never instead of it.

Related tools

  • Time-series storage sizing

    Tag count, sample rate and retention turned into terabytes, plus the edge buffer you need to survive a link outage of a given length.

  • Archive OCR project estimator

    Pages and scan quality turned into storage, processing hours and — the part that actually sets the schedule — the size of the human review queue.