Are Shelly devices safe on constant mains power?

[Tom001]

Hi everyone,

I’m considering using Shelly devices in my home, but I’m curious about their safety, especially since they are always connected to the mains power, which adds to my concern about durability and potential failure scenarios.

I’m particularly interested in understanding how likely it is for these devices to fail over time. What are common failure modes? Is there a chance of overheating or causing other damage that could lead to hazards like fires or toxic fumes, especially in places like bedrooms where safety is crucial?

I’d also like to know more about the built-in safety features. Can they effectively prevent worst-case scenarios, or is there a realistic risk to be concerned about?

This is a genuine question, as I want to ensure everything is as safe as possible before using them.

Thanks in advance for your insights!

Best,  
Tom

[terae]

I would say those are more safe than most of things you have always powered.

There's some safety things like most of them have 500mA fuse and VDR which do render the device just dead without more serious damages.

In case there's worst case, it's about wrong usage (like when you put it do completely wrong location which has too much moisture condense or a plenty of dust).

Worst case is arc inside the device, mostly when first power up and because of electrician made mistake with phases to control switch input from wrong one.

The plastic you have around these units, I have many times tested myself to be sure, those are safe. They melt, but to have them burn you need a lot of energy and anyway it's self extinguishing.

 

So. Place them in dry clean space. Not too much to ask when talking about electrical installations. But many doesn't pay enough attention.

[Heinz]

As someone who has built devices using Arduinos and relays, I understand firsthand the risks involved in creating and coding your systems. When building these devices yourself, the risk is higher, as even small mistakes—like incorrect placement of parts on a breadboard—can lead to serious failures. I've experienced this myself, having blown up devices due to simple errors in wiring or component placement.

While Shelly devices are pre-built and designed to meet safety and regulatory standards, they are still IoT devices, and proper care must be taken to ensure they are used correctly. This includes making sure they aren't overloaded or installed incorrectly. Even though these devices are built to be safer than DIY setups, mistakes during installation or improper use can still lead to hazards such as overheating or fire.

The reality is that any IoT device, regardless of the brand, carries some risk. No system is completely perfect. That's why it's important to regularly monitor these devices and ensure they are operating as intended. Whether it's a Shelly device or a custom-built solution, taking care to follow best practices for installation and maintenance is essential to minimizing risks and ensuring safe operation.

 

[Tom001]

Thank you for the two detailed responses and for taking this question serious!
I really appreciate the thorough explanations, and I’ll definitely take these points into account for my next steps.

On 24.9.2024 at 20:06, terae said:

I would say those are more safe than most of things you have always powered.

There's some safety things like most of them have 500mA fuse and VDR which do render the device just dead without more serious damages.

Good point, so this likely means that these two components are better than the integrated temperature monitoring since they interrupt the supply voltage as early as possible in the circuit, rather than just shutting off the outputs, which would probably be the response to overheating.
 

On 24.9.2024 at 20:06, terae said:

Worst case is arc inside the device, mostly when first power up and because of electrician made mistake with phases to control switch input from wrong one.

 

Ok, so that’s an installation error that would become immediately obvious. That’s not really my concern, as I find the gradual or spontaneous failure more worrisome.

How likely is an arc when the device is quietly sitting in the wall, just because a component fails due to age? Something that could happen even without active switching, but rather in a resting state.

 

13 hours ago, Heinz said:

As someone who has built devices using Arduinos and relays, I understand firsthand the risks involved in creating and coding your systems. When building these devices yourself, the risk is higher, as even small mistakes—like incorrect placement of parts on a breadboard—can lead to serious failures. I've experienced this myself, having blown up devices due to simple errors in wiring or component placement.

So you’re likely referring to fundamental design errors that could be avoided with experience. I would assume that with the widespread use of Shellys, such issues should not be present in their devices.
 

13 hours ago, Heinz said:

Even though these devices are built to be safer than DIY setups, mistakes during installation or improper use can still lead to hazards such as overheating or fire.

The reality is that any IoT device, regardless of the brand, carries some risk. No system is completely perfect. That's why it's important to regularly monitor these devices and ensure they are operating as intended. Whether it's a Shelly device or a custom-built solution, taking care to follow best practices for installation and maintenance is essential to minimizing risks and ensuring safe operation.

Yes, I completely agree. With Shelly devices, the challenge might be the ability to monitor their condition after installation, especially since they are built into the wall. I know there are monitoring systems like the temperature sensor, but checking the device after installation can be rather cumbersome.

I’ve been thinking about what could be improved inside a junction box for safety reasons. Metal inserts or something similar might make sense, although this would likely weaken the signal reception. On the other hand, I imagine there shouldn’t be much flammable material inside a recessed box in the event of a fire.

Aside from a fire, the fumes produced by smoldering are also a big concern, especially in areas like the bedroom where you might not wake up for a long time if this happens out of the blue.

In summary, I wouldn’t be concerned if a Shelly device were to fail after an acceptable usage time, like three years, as long as the failure is noticeable in a safe way, such as losing connection or having a relay issue. Their price is so reasonable compared to their benefits that I wouldn’t mind replacing them.

My main concerns really lie with how they fail, especially if the failure is spontaneous and occurs in a resting state or is associated with fire or smoke. I understand that every electrical device has a limited lifespan; the question is just how it ends.

Of course, this applies to every device. However, Shellys are permanently powered on and well hidden within the wall.

Thank you for the responses; I appreciate any further comments.

[borish]

I consider replacing my Bosch devices by Shellys, but I'm also concerned about the electrical safety of the Shellys. I have seen Amazon reviews that show Shellys that suddenly went up in fumes, like the one below. I have a shelly mini attached to a wodden pole and I don't feel comfortable when I see images like this.

[thgoebel]

Thanks for this picture (although it is not yours and you don’t know what was the reason for this fault)!

As much as I can see,

(a) the arcing was limited to the innards of the device - no melted case, no fire outside of the device, and

(b) arcing is located between the two SW inputs (at the right) of the Shelly plus 2PM, which let us assume it was a wrong wiring: Two different phases at both inputs.

[Heinz]

8 hours ago, Tom001 said:

Yes, I completely agree. With Shelly devices, the challenge might be the ability to monitor their condition after installation, especially since they are built into the wall. I know there are monitoring systems like the temperature sensor, but checking the device after installation can be rather cumbersome.

I’ve been thinking about what could be improved inside a junction box for safety reasons. Metal inserts or something similar might make sense, although this would likely weaken the signal reception. On the other hand, I imagine there shouldn’t be much flammable material inside a recessed box in the event of a fire.

Aside from a fire, the fumes produced by smoldering are also a big concern, especially in areas like the bedroom where you might not wake up for a long time if this happens out of the blue.

In summary, I wouldn’t be concerned if a Shelly device were to fail after an acceptable usage time, like three years, as long as the failure is noticeable in a safe way, such as losing connection or having a relay issue. Their price is so reasonable compared to their benefits that I wouldn’t mind replacing them.

My main concerns really lie with how they fail, especially if the failure is spontaneous and occurs in a resting state or is associated with fire or smoke. I understand that every electrical device has a limited lifespan; the question is just how it ends.

Of course, this applies to every device. However, Shellys are permanently powered on and well hidden within the wall.

Thank you for the responses; I appreciate any further comments.

When I say monitoring the device, I don’t mean constantly checking on it. Instead, it’s about periodically ensuring that the firmware is up-to-date.

In my experience, complete failure of an IoT device is usually due to how it's installed or used. I typically prefer Gen 2 or later versions because most bugs and issues have been addressed by then. This applies to all tech devices. If I get a Gen 1 device, I'm fully aware it's a first-generation product, and I’m prepared for potential problems. With Gen 2 or higher, I generally feel more confident about their reliability.

but of course its better to be on the safe side. 

[borish]

Here is another image. The user wrote he used the PM mini to meter a garden pump and it exploded after 2 weeks.

[Heinz]

Well with no context here we don't know anything of the type of motor and how long or what kind of load was on the Shelly PM Mini that caused it to explode after two weeks. 

it always helps knowing the reason on why something happened than to just post a picture with a destroyed device. 

[borish]

It was a 1300 W garden pump.

[terae]

3 hours ago, borish said:

Here is another image. The user wrote he used the PM mini to meter a garden pump and it exploded after 2 weeks.

Very rare failure and could be about moisture or something. Inductive loads are prone to cause arcing, but it should not go thru boad with these distances. How ever to have arc, there has to be massive overvoltage which causes that and it's less than second problem, nothing else will burn.

I have done many tests to cause arc and destroy some of Shelly devices....those are still safe.

[Tom001]

12 hours ago, borish said:

I consider replacing my Bosch devices by Shellys, but I'm also concerned about the electrical safety of the Shellys. I have seen Amazon reviews that show Shellys that suddenly went up in fumes, like the one below.

Are you considering Shelly devices because of their open platform or another feature compared to Bosch’s more integrated system?

11 hours ago, thgoebel said:

As much as I can see,

(a) the arcing was limited to the innards of the device - no melted case, no fire outside of the device, and

(b) arcing is located between the two SW inputs (at the right) of the Shelly plus 2PM, which let us assume it was a wrong wiring: Two different phases at both inputs.

Thanks for the analysis. Faults caused by incorrect wiring are acceptable, but spontaneous dangerous reactions are not.

10 hours ago, Heinz said:

In my experience, complete failure of an IoT device is usually due to how it's installed or used. I typically prefer Gen 2 or later versions because most bugs and issues have been addressed by then. This applies to all tech devices. If I get a Gen 1 device, I'm fully aware it's a first-generation product, and I’m prepared for potential problems. With Gen 2 or higher, I generally feel more confident about their reliability.

That’s an interesting point. I hope Shelly has improved their products based on past experiences.

7 hours ago, terae said:

Inductive loads are prone to cause arcing, but it should not go thru boad with these distances. How ever to have arc, there has to be massive overvoltage which causes that and it's less than second problem, nothing else will burn.

I have done many tests to cause arc and destroy some of Shelly devices....those are still safe.

So you're saying that the short duration of the arc isn't enough to set the device on fire, but the casing will turn black?

Thanks for mention your tests, that seems to be good. 

Can a statement be made about how intense the smoke development could be due to an arc? Will it come through the junction box into the room and pose a dangerous dose or concentration? I assume it depends on the amount of material and components involved.

I would use an RC snubber for shutters (inductive load) and ensure cleanliness and a dry environment. At the same time, I would want to keep the firmware up to date.

I don’t know what else I could do to minimize the risk. Mounting the Shelly outside wouldn’t work because it’s too humid there.

[terae]

13 hours ago, Tom001 said:

Are you considering Shelly devices because of their open platform or another feature compared to Bosch’s more integrated system?

Thanks for the analysis. Faults caused by incorrect wiring are acceptable, but spontaneous dangerous reactions are not.

That’s an interesting point. I hope Shelly has improved their products based on past experiences.

So you're saying that the short duration of the arc isn't enough to set the device on fire, but the casing will turn black?

Thanks for mention your tests, that seems to be good. 

Can a statement be made about how intense the smoke development could be due to an arc? Will it come through the junction box into the room and pose a dangerous dose or concentration? I assume it depends on the amount of material and components involved.

I would use an RC snubber for shutters (inductive load) and ensure cleanliness and a dry environment. At the same time, I would want to keep the firmware up to date.

I don’t know what else I could do to minimize the risk. Mounting the Shelly outside wouldn’t work because it’s too humid there.

 

Yes. There will be some metal seen inside when it does vapourize due high temperature and hit plastic case and due high pressure vapourizing metal causes the casing is almost every time opened up itself when arc hit. It is still safe when you follow normal electric installation guidelines.

Just don't install anything to where is saw dust and other dust and things just next to these devices and you'll be safe.

 

[Tom001]

On 27.9.2024 at 11:27, terae said:

 

Yes. There will be some metal seen inside when it does vapourize due high temperature and hit plastic case and due high pressure vapourizing metal causes the casing is almost every time opened up itself when arc hit. It is still safe when you follow normal electric installation guidelines.

Just don't install anything to where is saw dust and other dust and things just next to these devices and you'll be safe.

 

Thanks

If you've conducted many tests with arc faults and intentionally destroyed Shelly devices, can you describe how significant the smoke development was in such cases and whether there was a strong smell? Would a lot of smoke escape from an in-wall box into the room? Do you think a smoke detector would be triggered? Also, do you think there would be any health risks if someone were in the room during the event?

 

[terae]

1 hour ago, Tom001 said:

Thanks

If you've conducted many tests with arc faults and intentionally destroyed Shelly devices, can you describe how significant the smoke development was in such cases and whether there was a strong smell? Would a lot of smoke escape from an in-wall box into the room? Do you think a smoke detector would be triggered? Also, do you think there would be any health risks if someone were in the room during the event?

 

The smoke. Mostly nothing visible on air after explosion, possibly no one notice any visible smoke. Anyway you will notice bitter scent of burnt epoxy etc mixed with the odour of welding process. Yes you will notice it, it's something what people will notice easy with just small amounts in air. Next day your wife won't notice anything happened anyway. Otherwise I would have got some note to move thing or maybe myself too out of building...

The sound instead is something like old fashion toy gun detonator. Fast and loud bang. I don't like the sound, but the smell is almost good for a while. Call me weird if you wish. 😆