DC has a higher danger of arc faults than AC, and 48V is enough to sustain a short arc. But 48Vdc has less danger of either shock or arc faults than 120Vdc, and DC has a lower risk of shock than AC, because it still usually can't drive enough current through your skin to kill you. But you are correct that people can and do die from 48Vdc shocks, for example when their welding gloves are soaked with sweat or they're immersed in water.
You need thicker wires with 48V than with 120V or especially 240V, and the higher currents required at 48V can create more risk of fire from overheating conductors, for example in a wire that's too thin or a spring contact in an outlet that's worn out.
Perhaps you are implicitly comparing 48Vdc to 24Vdc, 12Vdc, or 6Vdc, rather than the 240Vac or 120Vac I was comparing it to. Those lower voltages do indeed need even thicker wires and pose even less risk of shock or arc faults than 48Vdc. I think 24V is the minimum to maintain an arc in air at atmospheric pressure, and it's really hard to get one started using a 24V supply; you need a high-frequency start circuit or a substantial amount of inductance in series.
48V is the maximum usually considered safe enough to not require compliance with any kind of electrical code; an expert electrician from North America once told me that it's in a category known to electricians as "bullshit wiring".
> DC has a lower risk of shock than AC, because it still usually can't drive enough current through your skin to kill you
What? Common wisdom is that DC shock risk is worse than AC, as DC makes your muscles clench up and so it's harder to let go of whatever you grabbed. That "120Vac" is actually 170V peak though, which does increase shock risk for equivalent power transferred (maybe this is what you meant?).
No, AC also makes your muscles clench up, and it's even worse. At a given RMS voltage, AC is both more painful and more dangerous as a source of shocks. Edison was kind of right about that.
It's complicated, though. If you're running a low-level current through your body for a long time, like in the neighborhood of a milliamp, DC is much worse because it migrates your electrolytes around. And if you're using metal electrodes it might be migrating the metal from the anode into your body. So TENS units are strictly AC, with no DC bias permitted.
I'm still trying to understand if you're saying AC itself has a greater shock risk, or if you're merely commenting on peak vs RMS. For safety terms I personally think of AC voltages as their peak voltage, because I consider it a bit ridiculous to model myself as a resistive load being heated - meaning I consider my house wiring here in the US as being 170V. And I would consider (a stiff source of) 170V DC more dangerous in all respects, but perhaps I am wrong?
You may be right that the higher peaks are the reason sinewave 120Vac RMS is a heftier shock risk. Or it might also be the fact of being AC, in the sense that a square wave inverter output with 120V peak and RMS voltage maybe is more dangerous than 120Vdc. I don't know if it is or not. I could make up reasons that it might be, like maybe it's harder for neurons to adapt to, or maybe electrolyte depletion near the electrodes increases the body's resistance, but I don't have any evidence.
There really are some shock risks that result from your body being heated as a resistive load, like internal burns, but those are not at the top of the list, especially at only 120Vac.
IIUC AC current causes your muscles to twitch. This means that can cause you to be unable to let go of e.g. a live wire, fall off a ladder, or disrupt your heartbeat.
You need thicker wires with 48V than with 120V or especially 240V, and the higher currents required at 48V can create more risk of fire from overheating conductors, for example in a wire that's too thin or a spring contact in an outlet that's worn out.
Perhaps you are implicitly comparing 48Vdc to 24Vdc, 12Vdc, or 6Vdc, rather than the 240Vac or 120Vac I was comparing it to. Those lower voltages do indeed need even thicker wires and pose even less risk of shock or arc faults than 48Vdc. I think 24V is the minimum to maintain an arc in air at atmospheric pressure, and it's really hard to get one started using a 24V supply; you need a high-frequency start circuit or a substantial amount of inductance in series.
48V is the maximum usually considered safe enough to not require compliance with any kind of electrical code; an expert electrician from North America once told me that it's in a category known to electricians as "bullshit wiring".