The Monsoon Office: A Power & Connectivity Continuity Plan for Small Teams

A waterlogged transformer in your area can pull power for four hours. A flooded fibre junction can drop your office internet at the same time. For a 15-person team running on cloud tools, that combination means no work — no email, no CRM, no calls — until both come back. This post is a continuity plan you can set up before the monsoon peaks: how to size a UPS, wire dual-ISP failover, pick offline-capable tools, and keep a remote-access fallback so a wet afternoon costs you minutes, not a day. ## What Is a Power & Connectivity Continuity Plan? A power and connectivity continuity plan is a small set of hardware and process choices that keep a team working through a power cut or an internet outage. For a small office, it has four parts: battery backup sized to bridge a multi-hour cut for the essentials, a second internet line that takes over automatically, tools that still function offline, and a way for staff to work from home if the office itself is unreachable. Set up once, it runs every monsoon. ## Why Plan for This Before the Rains Peak? The monsoon turns rare outages into weekly ones. Waterlogging trips transformers, lightning takes out exchanges, and flooded streets stop the lineman from reaching the fault. The India Meteorological Department's onset and progress updates make the timing roughly predictable each year — the [IMD monsoon tracker](https://mausam.imd.gov.in/) shows when your region is about to get hit. You want the kit installed and tested before the first big cell, not ordered the afternoon the power goes. The cost of not planning is concrete. A 15-person office where everyone bills time loses real money for every hour stalled. Worse, a half-saved invoice or an interrupted payment run can create errors that take longer to fix than the outage itself. The plan below is cheap insurance against both. ## The Four Layers of the Plan Think of continuity as four layers, each covering a different failure. You do not need all four on day one, but you should know which gap each one closes. ## How Do You Size a UPS for a Small Office? The most common mistake is buying a UPS for the wrong load. You do not need to power every desktop for four hours. You need to keep the network alive and let battery-powered laptops keep using it. Size for the router, the ONT or modem, the Wi-Fi access point, and maybe one shared monitor or the accountant's desktop. ## How Do You Set Up Dual-ISP Failover? Two internet lines only help if traffic moves between them without someone noticing and reconnecting cables. The right tool is a dual-WAN router that monitors both links and fails over automatically. ## Which Tools Keep Working Offline? Layer three is about choosing software that does not freeze the moment the link drops mid-failover. A short comparison of common categories:

Need	Offline-capable pick	What happens during an outage
Documents	Google Docs (offline mode on) or local LibreOffice	Edits save locally and sync when the link returns
Email	A desktop client (Outlook, Thunderbird) with cached mail	Read and draft offline; queued mail sends on reconnect
Accounting	Tally on a local machine	Fully functional offline; cloud-hosted Tally needs the network
Notes / tasks	Notion (limited offline), Obsidian (full offline)	Obsidian keeps working; Notion is read-mostly offline
Calls	Mobile phone as the fallback for client calls	VoIP drops with the internet; a charged phone does not

The lesson from the table: a fully cloud-only stack is the most exposed to outages. Keeping one offline-capable option per critical function means a brief drop during failover does not lose anyone's work. ## What Goes in the Remote Fallback? Layer four covers the case where the office building is the problem — flooded, evacuated, or with no power and no working inverter. The fallback is simple but must be written down before the day you need it. ## When Is This Overkill? Honesty matters here, because not every team needs the full kit. Three cases where you should scale down. If you are a fully remote team with no shared office, skip layers one and two entirely — your continuity problem is each person's home setup, so spend on staff power banks and a "if your line drops, tether and post in the channel" rule instead. If your office sits in an area with stable power and you have never lost the line for more than a few minutes, a single line-interactive UPS on the network and a backup 4G dongle in a drawer may be all you need. Do not buy a dual-WAN router and a tubular-battery inverter for a problem you do not have. And if your work is genuinely fine to pause — a design studio that bills by project, not by the hour, with no live customer support — a four-hour outage may simply be a long lunch. Match the spend to what an hour of downtime actually costs you. ## A Real Example: A 15-Person Architecture Studio in Pune A Pune architecture studio — 15 staff, heavy on cloud file storage and video calls with clients — lost three half-days to monsoon outages in one season. We set up the four layers in a day. Layer one: a 150Ah inverter with the router, ONT, switch, and two shared monitors on it, giving roughly four hours on the network load. Layer two: their existing fibre plus a second carrier's 5G router behind a MikroTik dual-WAN box, failing over in about 25 seconds in our test. Layer three: moved their active project notes to Obsidian and turned on Google Docs offline. Layer four: a one-page work-from-home plan with the VPN steps and the 8:30 am decision rule. The next cut hit on a Thursday afternoon. The lights went, the fibre dropped two minutes later, and the team kept editing and joined their 4 pm client call over 5G without a break. Manvi, who ran the setup and the failover test, noted that the only visible sign was the inverter's beep. The pattern is one we apply on larger reliability work through our web and app engineering engagements — the same uptime discipline behind client platforms like Radiant Finance — find each failure mode, close it cheaply, and test that the switch actually happens. ## Related Reading ## Frequently Asked Questions ### What size UPS or inverter does a 15-person office need? Size for the network, not every desktop. The router, modem, switch, and Wi-Fi together draw roughly 50 watts. A single 150Ah tubular-battery inverter stores around 1,800 watt-hours — easily four hours on the network plus spare for a laptop charger. Laptops run on their own batteries. ### Why use two different ISPs instead of two lines from one provider? A single flood or exchange fault often takes down everything from one provider at once. Two providers on different media — fibre plus 4G/5G — rarely fail together, so the pair gives real redundancy. Two lines from the same provider down the same conduit can drop simultaneously. ### How fast does dual-WAN failover happen? On a dual-WAN router with health checks, failover usually completes in around 30 seconds — the time it takes to notice the primary link stopped responding and route traffic to the backup. Active calls may drop and need a redial, but browsing and saved work resume on their own. ### Will my cloud tools work during the failover gap? Briefly, no — that 30-second gap can interrupt a live sync or call. That is why layer three matters: tools like Obsidian, a desktop email client, and Google Docs offline mode keep working through the gap and sync once the backup link is up. ### How much does the full kit cost for a small office? For a 15-person office, plan roughly ₹40,000–₹50,000 in hardware: an inverter with a tubular battery, a dual-WAN router, and a backup 4G/5G router. The recurring cost is the backup line's plan, often a low-rent metered SIM you only use during outages. ### Do remote and hybrid teams need this plan? Less of it. Fully remote teams skip the office power and dual-ISP layers and instead invest in each person's home resilience — a power bank, a tethering habit, and a clear "post in the channel if your line drops" rule. The remote-fallback layer becomes the whole plan. ### Can you set this up for us? Yes. We assess your office load and local outage pattern, spec the right kit, and configure dual-WAN failover plus offline-capable tooling — usually in a single day. We also write the remote-fallback page and run the first failover test with your team.