The VPN Topped My iPhone Battery Chart—but Reconnecting Was the Real Drain

By the time I reached the station, my iPhone battery had fallen to 24 percent. It had been above 60 that morning. I still needed the phone for my train ticket, directions after arrival, and a client message I had promised to answer during the journey. In Settings, the VPN sat at the top of the Battery list with hours of background activity. I blamed it, disconnected, and watched the client page immediately return to a UK access screen. Turning off the VPN protected the remaining percentage, but it also stopped me doing the work I needed to finish.

I had started leaving a VPN connected for longer periods after more UK websites introduced age checks. The Online Safety Act’s child-safety duties took effect on 25 July 2025, requiring covered services to use stronger age-assurance measures for certain material. VPN use rose sharply in the following year, although Ofcom did not attribute the entire increase to age checks.

That shift changed how people used VPNs. Something once opened for a single website was now running through commutes, cafés, station Wi-Fi and long periods on mobile data. Battery life was no longer a minor specification. It decided whether the phone would still work at the end of the journey.

My first explanation was simple: encryption must be consuming all the power.

The Battery screen appeared to prove it. The VPN had more background activity than almost any other app. But that chart was not measuring only the work performed by the VPN. Because other apps were sending traffic through its tunnel, iOS could attribute a large share of that network activity to the VPN itself. Independent electrical testing has shown that the percentage displayed beside a VPN can greatly overstate its direct power use.

The VPN was involved, but the chart was collapsing several different causes into one alarming number.

The better clue appeared below it: No Mobile Coverage and Low Signal.

Apple notes that an iPhone consumes more power when it has to search for cellular or Wi-Fi service. Weak reception keeps the radio working harder, while background apps continue trying to send and receive data. A VPN adds another connection that must stay alive while all of this is happening.

My commute was almost designed to expose the problem. The phone joined station Wi-Fi, failed to complete its sign-in page, returned to 5G, dropped to LTE inside the train, found another public network at the next station, and lost it again as the doors closed.

Each switch disturbed the VPN connection.

The established provider I was using had plenty in its favor. It had years of public history, a large support operation and applications for every device I owned. At home, on stable Wi-Fi, it rarely gave me a reason to open the app.

On the train, I kept opening it.

The status moved from connected to reconnecting. Messages waited. The client page stopped halfway through loading. I switched from the automatic location to the nearest server, thinking distance might be the problem. The connection returned, then disappeared after the next handoff between Wi-Fi and mobile data.

I changed protocols. That worked for part of the route, but the same cycle returned when the underlying network changed again. By then, the phone felt warm through the case, and I had spent more time watching the connection than answering the client.

A brief public user report described the same pattern during travel through poor coverage: severe battery drain appeared alongside an unstable connection rather than during ordinary use on strong Wi-Fi. That detail matched what was happening in front of me.

That changed the comparison.

I had been looking for a VPN that claimed to use less power while connected. What I needed was a VPN that spent less time losing the route, rebuilding it and forcing every waiting app to try again.

Turning the tunnel off was not an answer. The client page became inaccessible, and I lost the protection I wanted on station Wi-Fi. Low Power Mode reduced unrelated background work, but it could not stabilize the connection. Changing servers only restarted the same experiment in another country.

The next useful step was not another battery setting. It was a different way of handling the unstable network.

I opened OnlydogVPN and chose the preset for an unreliable connection. The interface did not ask me to decide among dozens of countries before it tried to solve the immediate problem. I connected and reopened the client page.

It loaded.

The more important result came a few minutes later. The train left another station, the public Wi-Fi disappeared, and the iPhone returned to mobile data. The page paused briefly, then continued instead of falling back to an error. The attachment I had been trying to send finished uploading.

I sent the reply and locked the screen.

That was the first moment on the journey when I stopped checking the VPN status.

The difference came from how the smaller app handled changes underneath the tunnel. It uses HTTP/3-based transport, built on QUIC, which is designed to keep connections moving when a device changes network paths. On a phone crossing between station Wi-Fi and cellular coverage, that matters more than maintaining the fastest possible result on one perfect connection.

The technology did not need a longer explanation. The observable result was enough: the network changed, the tunnel recovered, and the work continued.

After sending the client message, I opened a news page and noticed the app’s blocked-request counter rising. Advertising and tracking requests were being stopped before they completed.

That was a smaller benefit, but it belonged to the same battery problem. A modern page can continue making background requests after the useful text has already appeared. On weak mobile reception, every unnecessary request gives the phone more network work to perform.

The service reduced some of that work without requiring me to install a separate content blocker. Pages felt less busy, and the phone was no longer spending as much time loading elements I had not opened the site to see.

I could observe the blocked-request count and the resulting page behavior, but I could not inspect the service’s internal filtering rules. What I could verify was practical: the useful content arrived, fewer background elements followed it, and the phone remained cooler during the final part of the journey.

The steep battery decline did not repeat. I reached my stop with enough charge to open the ticket gate, follow walking directions and call the person meeting me.

That was the result I had needed when I first opened the Battery screen. I was not searching for a promise that a VPN would consume no power. Any active network service uses some. I needed the phone to remain useful until I arrived without giving up the access and protection that made me turn the VPN on in the first place.

The smaller service has a genuine limitation. It has a shorter public history and fewer independent reviews than the established provider. Someone choosing primarily on years of external scrutiny will find far more material about the older company.

But public history was not the scarce resource during this journey. Stable recovery was.

The established provider performed well when the phone stayed on one network. During the commute, it repeatedly made each signal change my problem. The smaller app recovered through the same kind of Wi-Fi-to-cellular transitions, while its request blocking reduced some of the background traffic competing for a weak connection.

The iPhone Battery screen had encouraged me to judge the app with the largest percentage beside its name. That was too crude. A VPN sits in the path of every connected app, so its row can become a summary of traffic passing through it rather than a clean measure of its own cost.

A more useful test began when the signal weakened. Did the tunnel reconnect before the page failed? Did the upload continue? Did the phone stay cool enough—and charged enough—to finish the journey?

For this iPhone problem, efficient recovery mattered more than an impressive low-power claim. I did not need a VPN that looked innocent in the battery chart while repeatedly rebuilding its connection. I needed the one that let me stop checking the percentage before I reached the ticket barrier.