I’m toying with the idea of a car engine + generator for power backup

[Jester]

I though about the radial loading, I would support the shaft with bearings on either end, that should take care of that., I doubt it would snap the crankshaft, but it would likely do a number on the rear bearing and seal.

I was in my car today and made a point of listening to how loud it was at say 1500, 1800 3600 RPM, pretty darn quiet at 1500-1800 RPM even while accelerating, not a completely fair comparison though as my daily driver is a 3.0 litre V6.

I also tried to find HP vs. RPM graphs for 1.5l engines, no shortage for “hopped up” engines, I did not find a graph for a totally stock Toyota or Honda, but I’m guessing it will make plenty of power at 1500RPM to comfortably power the house. I’m really not concerned with consumer reliability vs. industrial, if people put 400km+ on these engines without a rebuild I’m guessing running for a few days at 1500RPM will be nothing.

it adds up quickly, say a 5000km oil change interval, guessing an average speed of 50km/h, that's only 100hours

Well on my modern Volvo the service intervals are 25 000km. This is about how much i drive in a year, so it gets a yearly oil change. Most of this time is spent going 130km/h on a highway. Googling around it seams to take about 40 HP to maintain this speed in an average car. So on the typical commute to work it is continually producing 40 HP for about 30 minutes. This ads up to about 240 hours of running at 40 HP between oil changes.

So it looks like a typical car engine should be able to produce about 30kW for 10 days straight without service. In the process it would burn about 1500 liters of fuel(or about 1000$ to 4000$ worth of fuel depending on where you live). Tho 30kW is a lot of power so a house probably won't use nearly that much 24/7

Exactly, the 20-30kW is a pie in the sky number based on:
If I do this I may as well do it well enough once, so if I want to use worst case power; electric range, furnace etc. I don’t need to think about what breakers do I need to turn on and off now or worry about exceeding the current limit while two motors decide to start simultaneously. I want to simply start the generator, transfer over and forget about it while the fuel lasts or until the power comes back on. In reality will probably average 2kW. The chances of a 10 day outage are really really low, worst case we have had in the last 60 years was about 2-3 days.

It would also allow us to send a little power to our neighbours to keep their heat on in a pinch.

[brucehoult]

So it looks like a typical car engine should be able to produce about 30kW for 10 days straight without service. In the process it would burn about 1500 liters of fuel(or about 1000$ to 4000$ worth of fuel depending on where you live). Tho 30kW is a lot of power so a house probably won't use nearly that much 24/7

There was this interesting example, almost 33 years ago in early 1989: take a standard family car (1st gen Subaru Legacy with 2.0 turbo engine) chosen by the FIA at random from the production line and do 100,000 km in it as quickly as possible.

The result? 447 hours, 44 minutes, and 9.887 seconds (about 18.6 days), at an average speed of 223.345 km/h (138.78 mph)

They refuelled and changed drivers every two hours, and tyres every 96 hours (about 21500 km). I know they also did oil changes, but I don't know how often.

As the average speed includes all the stops, the actual speed when moving was normally around 145 mph.

That will have been putting out a bit more than 30 kW :-) That engine was rated at about 160 kW, and will have been using most of that.

[james_s]

There was this interesting example, almost 33 years ago in early 1989: take a standard family car (1st gen Subaru Legacy with 2.0 turbo engine) chosen by the FIA at random from the production line and do 100,000 km in it as quickly as possible.

The result? 447 hours, 44 minutes, and 9.887 seconds (about 18.6 days), at an average speed of 223.345 km/h (138.78 mph)

They refuelled and changed drivers every two hours, and tyres every 96 hours (about 21500 km). I know they also did oil changes, but I don't know how often.

As the average speed includes all the stops, the actual speed when moving was normally around 145 mph.

That's a shockingly high average speed. Most cars of that era are incapable of reaching 145 mph due to the gearing regardless of available power, the vast majority of those that can will struggle to reach it and require a long flat straightaway to get there. I'm very surprised that a little 2.0L Subaru could manage to sustain a nearly 140mph average for more than an hour or two without something blowing up.

[brucehoult]

I'm very surprised that a little 2.0L Subaru could manage to sustain a nearly 140mph average for more than an hour or two without something blowing up.

I expect they surprised a lot of people. I believe this record category was previously held by Mercedes.

They started with three cars. All three finished the 100,000 km at the same time, side by side.

[langwadt]

There was this interesting example, almost 33 years ago in early 1989: take a standard family car (1st gen Subaru Legacy with 2.0 turbo engine) chosen by the FIA at random from the production line and do 100,000 km in it as quickly as possible.

The result? 447 hours, 44 minutes, and 9.887 seconds (about 18.6 days), at an average speed of 223.345 km/h (138.78 mph)

They refuelled and changed drivers every two hours, and tyres every 96 hours (about 21500 km). I know they also did oil changes, but I don't know how often.

As the average speed includes all the stops, the actual speed when moving was normally around 145 mph.

That's a shockingly high average speed. Most cars of that era are incapable of reaching 145 mph due to the gearing regardless of available power, the vast majority of those that can will struggle to reach it and require a long flat straightaway to get there. I'm very surprised that a little 2.0L Subaru could manage to sustain a nearly 140mph average for more than an hour or two without something blowing up.

specs say 200hp and 230km/h topspeed

[brucehoult]

specs say 200hp and 230km/h topspeed

Which specs?

At 3:10 this video shows specs of 220 PS @6400 RPM, 27.5 kgf-m @4000 RPM.

US models often have larger engines (e.g. 2.2 in 1st gen Legacy), lower power output, and weigh more than JDM (and often Aus/NZ) models.

For example my Japanese 2008 Outback 2.5XT is rated at 265 HP @5600 (250 N-m, 35.7 kgf-m, 258 lbf-ft @2400). The model with the same name in the US was rated at 250 HP.

Here's Subaru's press release at the time https://www.subaru.co.jp/news/archives/08_04_06/08_05_08_02.html

My car is doing 2100 RPM at 100 km/h, so in theory is geared for 266 km/h at RPM for maximum power. It's got quite a lot more power than that 1989 model, and I'd say better aerodynamics, so it might be possible.

I was very happy to pick it up in excellent condition for US$5900 with 87,000 km on the clock and I've done another 40,000 so far.

https://pbs.twimg.com/media/FFBMxKbVUAENFD9.jpg

Having dynamic cruise control, lane departure warning, automatic pre-collision braking etc in a 2008 car is really great. It's even got quite reasonable fuel consumption :-)

https://pbs.twimg.com/media/FFBMxKbVUAENFD9?format=jpg&name=large

(every tank with worse than 10 l/100km or 24 USMPG was towing something)

[BrokenYugo]

Another concern with putting a car engine in generator service comes to mind, oil capacity and control. Most auto engines have small sumps, no low oil level detection/shutdown, and piston rings oriented more towards minimal drag than perfect oil control, even when they aren't gunked up from insufficient oil changes or puttering around constantly. Not unusual for it to be perfectly acceptable for such an engine, in warranty, to blow though a whole sump load of oil (or more!) within the recommended oil change interval, and put a rod through the block if the owner doesn't keep up with it.

[langwadt]

Another concern with putting a car engine in generator service comes to mind, oil capacity and control. Most auto engines have small sumps, no low oil level detection/shutdown, and piston rings oriented more towards minimal drag than perfect oil control, even when they aren't gunked up from insufficient oil changes or puttering around constantly. Not unusual for it to be perfectly acceptable for such an engine, in warranty, to blow though a whole sump load of oil (or more!) within the recommended oil change interval, and put a rod through the block if the owner doesn't keep up with it.

doesn't many newish cars with long service intervals have oil level and quality sensors? if modern cars didn't have pretty good oil control they would fail emissions miserably 

[BrokenYugo]

We're talking about cheap junkyard stuff here though, late 90s-00s. And the latest still have consumption issues here and there. Seems it was actually a fairly narrow window between good hard rings and bores that would keep an engine tight for 200k+ (early 90s?) and everybody getting greedy with ring tension and whatnot.

You would think it'd trash the cats too fast, but even now the threshold for "you get a free motor" is usually somewhere around 1000 mile/qt.

[langwadt]

We're talking about cheap junkyard stuff here though, late 90s-00s. And the latest still have consumption issues here and there. Seems it was actually a fairly narrow window between good hard rings and bores that would keep an engine tight for 200k+ (early 90s?) and everybody getting greedy with ring tension and whatnot.

You would think it'd trash the cats too fast, but even now the threshold for "you get a free motor" is usually somewhere around 1000 mile/qt.

do junkyards still have stuff that is +20 years old?

[brucehoult]

do junkyards still have stuff that is +20 years old?

Pretty sure it's still easy to get parts for any reasonably popular model from the 90s here in New Zealand. I only just sold my 1997 Outback 2 1/2 years ago, and that was only because I was moving to the USA and thought I might not be back for 5 or 10 years (that didn't work out) so no point storing it and coming back to a 30+ year old car.

My new car is 13 years old and I see no reason I won't still be driving it in a decade from now -- I'm more likely to die than it is.

[james_s]

do junkyards still have stuff that is +20 years old?

They do around here, occasionally quite a bit older. It really bugs me when I see a reasonably nice straight older car in the yard though, usually with some fairly easily repairable fault, sometimes nothing wrong at all. There's a "cash for junk cars" thing around here that advertises pretty heavily, a lot of people think the cars are going to charity but in reality they go to their huge chain of salvage yards to get scrapped.

[bdunham7]

do junkyards still have stuff that is +20 years old?

Here they haven't gotten the 20-year old cars yet except for wrecks--the rest are still out on the road.  I see 50+ year old cars out on the road regularly and even 100-year old cars on occasion.  Junkyards are full of stuff going back to the 80's, but probably only the stuff that still has some demand.

[Berni]

There was this interesting example, almost 33 years ago in early 1989: take a standard family car (1st gen Subaru Legacy with 2.0 turbo engine) chosen by the FIA at random from the production line and do 100,000 km in it as quickly as possible.

The result? 447 hours, 44 minutes, and 9.887 seconds (about 18.6 days), at an average speed of 223.345 km/h (138.78 mph)

That is indeed rather impressive.

I was mostly talking about average numbers that most car engines should easily meet since its part of what these cars do in regular daily use while abiding by the manufacturers recommended service interval. Most engines will likely do 4x more miles than the service interval without a oil change and still run fine. It's just not a good idea to do on an engine you care about (especially modern ones) since one important oil passage clogging up with gunk is all it takes to seriously wreck the engine.

Another good example of how long an engine can run is the record attempt for the longest duration flight where two guys flew circles for 64 days in a little Cesna. Tho aircraft engines are designed for running at high throttle for long periods, but this put 1500 hours on that poor engine, well over the short service intervals for these engines. To do this they not only did in flight refueling but also inflight oil changes by running hoses to the engine and flushing new oil trough a few times. Eventually the engine got gunked up with enough carbon to make it loose power and they had to land.

doesn't many newish cars with long service intervals have oil level and quality sensors? if modern cars didn't have pretty good oil control they would fail emissions miserably 

Even very old engines typically have oil pressure lights, this comes on when the oil pump starts sucking air. Slightly more modern ones typically have oil level sensors too, very nice to give you a warning before shit hits the fan. Especially since modern engines indeed drink a bit of oil even new.(Optimized for low drag and using low viscosity oils)

However i don't think car engines turn off when they run out of oil. Typically they will throw up plenty of warnings on the dash, but keep on running, presumably until they seize up or break.

[james_s]

Even very old engines typically have oil pressure lights, this comes on when the oil pump starts sucking air. Slightly more modern ones typically have oil level sensors too, very nice to give you a warning before shit hits the fan. Especially since modern engines indeed drink a bit of oil even new.(Optimized for low drag and using low viscosity oils)

However i don't think car engines turn off when they run out of oil. Typically they will throw up plenty of warnings on the dash, but keep on running, presumably until they seize up or break.

They don't shut off, I've known a couple different idiots who ran their cars out of oil. A friend of mine used to be a professional mechanic and a girl he knew told him the "low oil light" on her car had been on for 3 days and then it wouldn't start, this was of course the low oil pressure light and the engine was completely destroyed.

I've never understood why there is not an obnoxious buzzer for low oil pressure and high coolant temperature. Lots of cars had a buzzer like that if you don't have the seatbelt on but I've never seen anything more than a warning light and gauge for lack of oil pressure.

[Jester]

IMO the odds of those three Subaru’s being truly stock is zero!

You can be sure those cars were disassembled blueprinted ( probably tweaked) and reassembled by a specialist with years of experience.

I recently read an article about an endurance test Goodyear did To promote their new tire of the day, I think they raced a Mustang or Shelby around Indy or some similar oval track at high speeds like that for an extended period.

Apparently the entire power train, cooling system etc. was tweaked, better bearings, cooled transmission and differential etc.

I’m not saying you could not put a bone stock production car through this this and not have it fail, but plenty of them would and that would pretty much defeat the point of the advertising gimmick, so you invest the time and money upfront to minimize the odds of a failure.

I don’t know what kind of cars you guys are driving that burn through a pan of oil between changes, but on every Toyota and Lexus that we have had in the family about 10 over the years the oil level does not go down any noticeable amount between changes, in fact my Lexus for the first 150,000 km or so the oil still looks clean clearish yellow not brown at oil change time. I have a friend with a Corolla with 450,000km fluid and plug changes I think one new water pump. I have high confidence that even a high mileage engine could survive at 1500 RPM producing 10-20 kW for a good long time.

My wife had an outback, must have been really unlucky engine failed twice under warranty and it did consume oil between changes typically at least a litre. Shortly after warranty it started knocking again, traded it in for a Sienna, will never buy a Subaru again. I have had a lot of really reliable cars over the years and only two duds, that required more repairs than all the others combined, last place was the Pinto and runner up was the outback.

[langwadt]

IMO the odds of those three Subaru’s being truly stock is zero!

You can be sure those cars were disassembled blueprinted ( probably tweaked) and reassembled by a specialist with years of experience.

I recently read an article about an endurance test Goodyear did To promote their new tire of the day, I think they raced a Mustang or Shelby around Indy or some similar oval track at high speeds like that for an extended period.

Apparently the entire power train, cooling system etc. was tweaked, better bearings, cooled transmission and differential etc.

I’m not saying you could not put a bone stock production car through this this and not have it fail, but plenty of them would and that would pretty much defeat the point of the advertising gimmick, so you invest the time and money upfront to minimize the odds of a failure.

I don’t know what kind of cars you guys are driving that burn through a pan of oil between changes, but on every Toyota and Lexus that we have had in the family about 10 over the years the oil level does not go down any noticeable amount between changes, in fact my Lexus for the first 150,000 km or so the oil still looks clean clearish yellow not brown at oil change time. I have a friend with a Corolla with 450,000km fluid and plug changes I think one new water pump. I have high confidence that even a high mileage engine could put along at 1500 RPM producing 10-20 kW for a good long time.

My wife had an outback, must have been really unlucky engine failed twice under warranty and it did consume oil between changes typically at least a litre. Shortly after warranty it started knocking again, traded it in for a Sienna, will never buy a Subaru again.

"Toyota and Lexus" thats' why, expect for certain years of Avensis with high fuel comsumption that was later fixed, afaik they are pretty much bulletproof
 

[AaronD]

I don’t know what kind of cars you guys are driving that burn through a pan of oil between changes, but on every Toyota and Lexus that we have had in the family about 10 over the years the oil level does not go down any noticeable amount between changes, in fact my Lexus for the first 150,000 km or so the oil still looks clean clearish yellow not brown at oil change time. I have a friend with a Corolla with 450,000km fluid and plug changes I think one new water pump. I have high confidence that even a high mileage engine could survive at 1500 RPM producing 10-20 kW for a good long time.

My Jeep takes 6 quarts of fossil 10W-30, and I typically get 5 quarts out after 5,000 miles.  Still more than plenty.

I'm having a hard time finding pure fossil oil though, so what's in it now is a "synthetic blend".  The engine is new enough that (according to what I've heard through the grapevine) I could probably switch to full synthetic and still be okay, which is supposed to come with a longer service interval.  That would make the synthetic cheaper overall, despite being more expensive to buy the same volume of it, but does it really work like that for an engine that is designed for fossil oil?

[james_s]

There is absolutely no problem with running synthetic oil in any engine, it is superior in every way except cost to conventional oil, it's all made from the same crude oil stock anyway. The *only* time when it may be preferable to run conventional oil is during break-in of a newly rebuilt engine when you want parts to wear slightly to seat the rings and such and in that case the superior lubricity of synthetic oil can work against you.

I've run synthetic oil in my turbo cars for as long as I've had them, the oldest is a 1984 which was made back when I don't think synthetic oil was around. Never had any issues at all related to the oil.

[AaronD]

That's good to know.  I've also heard something about conventional oil being a natural leak stopper as well, because the sludge that it breaks down into, tends to get stuck there.  When you switch to synthetic, all that sludge slowly comes out of its little crevices, and you start getting leaks that you never had before.  Is that true?

[bdunham7]

There is absolutely no problem with running synthetic oil in any engine, it is superior in every way except cost to conventional oil, it's all made from the same crude oil stock anyway.

That's a bit of an overgeneralization.  Although mostly true nowadays, there are plenty of past examples that would refute each of those statements.  There's quite a bit of variation in synthetic oils and it it takes a PhD in tribology to comprehend them all.  And it is perfectly possible to synthesize a terrible quality motor oil,  a feat which was accomplished in earlier years.

[Jester]

There is absolutely no problem with running synthetic oil in any engine, it is superior in every way except cost to conventional oil, it's all made from the same crude oil stock anyway. The *only* time when it may be preferable to run conventional oil is during break-in of a newly rebuilt engine when you want parts to wear slightly to seat the rings and such and in that case the superior lubricity of synthetic oil can work against you.

I've run synthetic oil in my turbo cars for as long as I've had them, the oldest is a 1984 which was made back when I don't think synthetic oil was around. Never had any issues at all related to the oil.

Synthetic oil has been around for a while, I recall reading that the finer machinery used in the panzer tanks was jamming up in the frigid temperatures of Russia, while the cruder Russian tanks had no problems with the cold temperatures because of the loose tolerances. Solution synthetic oil.

I use conventional oil in my daily driver, synthetic in my hot rod.

[bdunham7]

That's good to know.  I've also heard something about conventional oil being a natural leak stopper as well, because the sludge that it breaks down into, tends to get stuck there.  When you switch to synthetic, all that sludge slowly comes out of its little crevices, and you start getting leaks that you never had before.  Is that true?

Group iV PAO-based synthetics are actually not the best lubricants and don't have the same effect on seals that mineral-based oils do, so they need to be properly blended with additives (esters) that could mimic those functions.  Some early synthetics were not all that well engineered and had some of the issues you mentioned.  I'm pretty sure any reputable manufacturer nowadays has solved those issues by now, so you only need to make sure that the oil you buy has the specific ratings required by your vehicle's manufacturer to be at least minimally sure of having the right stuff.  I wouldn't recommend extending the service intervals though, so I'm not sure you will get any advantages with so-called full-synthetic product over a good quality blend.

[james_s]

That's a bit of an overgeneralization.  Although mostly true nowadays, there are plenty of past examples that would refute each of those statements.  There's quite a bit of variation in synthetic oils and it it takes a PhD in tribology to comprehend them all.  And it is perfectly possible to synthesize a terrible quality motor oil,  a feat which was accomplished in earlier years.

Of course it's possible to synthesize a terrible oil, but that's a good product vs terrible product comparison, not a synthetic vs conventional oil comparison. I am not aware of any engine that will not perform properly or that will be damaged somehow by using a synthetic oil of the proper viscosity. Engine oils are tested and certified to meet standards, I forget what it's called, but any oil that meets the standard and has the correct ratings for the engine should be fine. There have been numerous cars in my family with over 200k miles on them, one that my mother has owned since new has over 340k on it and I had a daily driver that had 330k when somebody rear ended me and totaled it. I've never done anything special with the oil other than making sure there was always oil in them and changing it on a reasonable basis.

[james_s]

That's good to know.  I've also heard something about conventional oil being a natural leak stopper as well, because the sludge that it breaks down into, tends to get stuck there.  When you switch to synthetic, all that sludge slowly comes out of its little crevices, and you start getting leaks that you never had before.  Is that true?

I'm of the opinion that if something starts leaking, it probably was time to replace the seal/gasket/whatever anyway. I know that after years of running synthetic when I remove the valve cover everything looks much cleaner than similar engines that have always run conventional oil so it certainly does seem to dissolve a lot of crud that wasn't getting dissolved before. I personally have not had issues caused by that but I suppose anything is possible. If you are curious about change intervals, there are companies you can send a sample of the used oil to and they will analyze it and send you a report showing the lubricity, state of the various additives, acidic content, quantities of various metals and other wear products from the engine and that can give you a good idea of the overall condition of the oil. If the oil is still in good shape when you change it that's a good indication that you can safely extend the change interval. Always use a good quality filter too.