Cycling UK Forum

Here's a reputable source on global incidents involving BESS (Battery Energy Storage Systems) events - https://storagewiki.epri.com/index.php/ ... t_Database

Here's a screenshot of an overview of the locations and frequency over the last 10 or so years from that source. This article https://www.researchgate.net/profile/Wa ... ystems.pdf describes the potential hazards of BESS. It's a sobering read and urges the UK government to apply higher standards of regulation to their planning and operation. I've highlighted a few excerpts from the executive summary below.

Despite storing electrochemical energy of many hundreds of tons of TNT equivalent, and several times the energy released in the August 2020 Beirut explosion, these BESS are regarded as “articles” by the Health and Safety Executive (HSE), in defiance of the Control of Major Accident Hazards Regulations (COMAH) 2015, intended to safeguard public health, property and the environment. The HSE currently makes no representations on BESS to Planning Examinations.

Li-ion batteries can fail by “thermal runaway” where overheating in a single faulty cell can propagate to neighbours with energy releases popularly known as “battery fires”. These are not strictly “fires” at all, requiring no oxygen to propagate. They are uncontrollable except by
extravagant water cooling. They evolve toxic gases such as Hydrogen Fluoride (HF) and highly inflammable gases including Hydrogen (H2), Methane (CH4), Ethylene (C2H4) and Carbon Monoxide (CO). These in turn may cause further explosions or fires upon ignition. The chemical energy then released can be up to 20 times the stored electrochemical energy. Acute Toxic gases and Inflammable Gases are “dangerous substances” controlled by COMAH 2015.

The scale of Li-ion BESS energy storage envisioned at “mega scale” energy farms is unprecedented and requires urgent review. The explosion potential and the lack of engineering standards to prevent thermal runaway may put control of “battery fires” beyond the knowledge, experience and capabilities of local Fire and Rescue Services. BESS present special hazards to fire-fighters; four sustained life-limiting injuries in the Arizona incident.

We can expect many 'mega scale' BESS in the future if we are to have adequate back-up during extended periods of low renewable generation. The Liverpool BESS fire (Sept 2020) listed in the database was in a pretty small BESS (10MWh at 20MW; so equivalent to half an hour's storage at its nominal power output). The largest BESS in the database stored over a 100 times more electrochemical energy (1200MWh), and with 15 times the nominal output (just 4 hours storage at that output). We'll need to get used to them!

I have a similar requirement to Tangled Metal, and I have recently bought a Montana 750i. The device comes into its own as Hiking GPS with its large, bright and high-resolution screen. I had found planning hiking routes in the field, as it were, with an Edge 1030 increasingly unsatisfactory because the screen isn't quite clear enough, and is just a little too small to see more distant paths etc. on 1:50k and 1:25k OS mapping. Of course, the 1030 is a pretty big screen as cycling GPSs go, and is it really good for most cycling - probably the best 'cycling' GPS I've owned. Also, I use the 750i with marine charts for sea kayaking, and I wouldn't trust the Edge to be sufficiently waterproof for this activity.

I was a long-time user of the Montana 600, and I share Mick F's views on it, ... that is when it is working properly. The problem was it would randomly shutdown whilst following planned routes, sometimes many times a day; it would always re-start quickly though once I'd noticed and switched it on again, so it was not a serious problem. I probably had 5 or 6 devices in total, each one swapped FoC by Garmin. However, if I wasn't following a route, then it was much more reliable, which meant it was good for hiking, since I rarely use a pre-loaded hiking route. Needless to say, it wasn't great for audaxing if you needed a complete .gpx track file of you ride. That said, my Montana had fallen out use for hiking in preference for the Edge 1030, despite it not having an electronic compass.

The main drawback with the 750i is its size; it is quite a beast, and noticeably larger than the Montana 600. However, the excellent and larger screen makes route planning on the hoof almost as good as paper map, although I often print an OS map of the area too using Basecamp and the map stored on the device anyway as back up. The problem with smart phones, in my experience, and not withstanding the more limited battery life, is the screens can be very hard to read in bright sunlight, although I expect careful selection of a particular phone might overcome this problem to a degree. The Montana screens, whether 600 or 750i, are particularly clear in direct sunlight, which is essential for a decent outdoor GPS in my opinion.

Battery life for modern GPSs is very good, although the Montana 600 isn't up to the standard of the 1030 or 750i. I usually have a generic USB cache battery with me too, but it's much easier to swap out batteries, so I carry a spare Li-ion battery (not for the 1030, of course) too. I rarely use AA batteries, but that was an option with the Montana 600.

I also use my Montana 750i on my touring bike. I'm not a weightweenie anyway, but the device plus the mount is quite a size. I adapted an AMPS Rugged charging mount for motorcycling, and improvised with a slightly lower profile (and lighter, for all it matters!) attachment to the bike and mashed up a USB plug on the cable to charge it via my hub generator charging system. It's a pretty clean and weatherproof solution, and I've been very impressed with how it keeps the 750i fully charged even on hilly, off-road days.

In terms of connectivity, the 750i is pretty comprehensive, but it won't handle some cycling ANT+ sensors. However, I have found an IQ app for the 750i that now enables it to connect via ANT+ to a power meter. I have a gripe about its connectivity to the Garmin Connect and Garmin Explore apps though. These two apps don't really seem talk to each other currently, and the 750i will send information to both, but it will only received information from Garmin Explore. I find Garmin Connect is better for planning cycling routes, but getting the route wirelessly onto the 750i is a problem; it's easy with access to a computer, of course. Instead, I have found Komoot is an intermediary, but surely I shouldn't need to go via a third party app. Komoot has an IQ app for Garmin devices and is a good means of planning routes anyway. I used to use BikeHike, which was good for planning routes for the Montana 600, but I've not used it much recently.

I've gone on enough, and it's probably only Tangled Metal reading this far, sorry.

Chris Stark, who's CEO of the Climate Change Committee, has claimed that it is 'better' to ride an electric bike than a conventional bike. You can hear his claim at ~31 minutes in the 16 June 2021 edition of the BBC Newcast podcast.

You can find it via BBC Sounds or using this link here: https://www.bbc.co.uk/sounds/play/p09lphx0. [I apologise for the programme title not complying with the forum rules!]

Sounds fanciful to me.

This is an interesting website I came across this morning: https://carbonintensity.org.uk/. It aims to track the 'carbon intensity' of the UK power generation.

This is the map of the regional 'carbon intensity' this morning (Monday 7th December) at 9.30am. What's noticeable is the green 'low' areas are low because of the predominance of nuclear generation (>50% in the three green regions). The two 'moderate' yellow areas are principally because of hydro generation (i.e. Northern Scotland) or imported electricity, which is presumably nuclear-generated electricity from France (i.e. SE England).

Just for the record, it's a cloudy morning with little wind across much of the UK today. Solar generation is virtually nil, and wind generation nationwide this morning was ~5% (which is up from ~0.8% yesterday, which was even less windy than today). Also the regions with the highest contribution from wind generation at ~10.5% and ~8.5% this morning were East England and Yorkshire respectively, perhaps surprisingly low contributions considering the proximity of the very large offshore windfarms in the North Sea (both regions indicated as 'high' carbon intensity). East England has Sizewell B representing ~21% of the regional generation, and Yorkshire 'benefits' from a large biomass contribution of ~31%, which I imagine is mainly the huge Drax power station burning wood imported from North America. South West England's was slightly more than 98% gas-fired generation, from which I assume the Hinkley B nuclear power station is currently offline.

Another interesting point is that this website recommended that users unplug and switch off appliances and devices between 8-10am this morning to reduce the overall demand for electricity when the carbon intensity was high.

Finally, just returning to the original subject of this thread, perhaps the adoption of a fleet of small modular reactors spread around the country will even out and lower the 'carbon intensity' of the regions, particularly in winter when electricity demand is high.

pete75 wrote:Strange thing is Germany is getting rid of nuclear power altogether yet here we have the British subsidiary of a German company proposing mini nuclear plants.

You are confusing Rolls-Royce Motor Cars Ltd, wholly owned by BMW, with Rolls-Royce plc, the aero-engine and nuclear business. They haven't been part of the same organisation for ~40 years.

You may be surprised that coal still has the largest share of German power generation, and coal-fired power stations are due to be fully phased out by 2038 (see https://www.iea.org/reports/germany-2020).

If anyone is interested in an SPD pedal power meter, then the Assioma power meter offers an unexpected opportunity. The Assioma pedals are identical to an Xpedo road pedal, which shares the same spindle with several other Xpedo pedals, including some Xpedo SPD-compatible pedals. [See here: https://jnyyz.wordpress.com/2020/07/24/ ... edal-hack/.] It is a simple job to swap over the pedal bodies to build an SPD-pedal power meter for little more than half the cost of the new SRM pedal power meter. It was fortunate that I already owned a pair of compatible Xpedo pedals. A little care is needed to check for interference between the shoe's tread and the pod containing the power meter electronics etc., but I've not found it to be a problem so far, even with my heavy duty Lake winter boots. I'm sure a judicious use of a craft knife would resolve any problem if one were to arise.

I can report good experience with the Assioma pedal power meter too. The battery life is good - claimed to be 50 hours (vs. 30 hours claimed for the SRM pedals, I recall). Charging is no more difficult than charging your phone or GPS, and a dual-socket USB battery pack and the supplied cables would do the job in a hour or so (or a single-socket USB battery by charging one pedal at a time). In fact, you could use the same micro-USB cable as you use for your phone/GPS, and just carry the lightweight proprietary magnetic adapter that clips on to the pedal.

I'm looking forward to fitting two pairs on the tandem, and then we'll see who's working the hardest!

I don't believe the recent Which? diesel emissions tests have been reported much in the wider media. So, here's the graphical representation of some of the results:


Source: http://press.which.co.uk/whichpressrele ... variation/

Two points are worth highlighting from the accompanying notes:

1. "Which? has tested averages for VW Group cars (VW, Audi, Seat and Skoda), and they are some of lowest measured NOx averages. However, the Euro 5 diesel cars tested are part of the ongoing VW emissions investigation, and so a question mark remains over the results from these cars."

2. "For more information on our tests and how they differ from official EC tests, go to: http://www.which.co.uk/reviews/cars/art ... -emissions". This Which? testing protocol includes: "How we stop manufacturers from cheating our tests: There is a risk that cars (from any manufacturer, not just VW) could use similar means to detect they are in a lab testing environment and switch to a more economical running mode. To stop any manufacturer confusing our tests, whenever we are suspicious of the NOx emissions we record in the lab, we now run additional test cycles using a Portable Emissions Measuring System (PEMS)."

It is not clear whether Which? invoked use of its PEMS testing in these tests in order to 'defeat' the so-called 'defeat devices', such as those reported in VW Euro 5 diesel cars.

I don't think this has been posted in this thread so far - I apologise if I've missed it. Here's a similar plot and looking back to 1970.


Source: https://www.gov.uk/government/uploads/s ... _final.pdf

Psamathe wrote:

fc101 wrote:...
Have a look at the book, and you'll see it is very relevant to this debate - probably more than half the book is devoted the UK renewable electricity generation potential. Also, for example, some of the foregoing discussion in this thread has been about transport, which will increasingly be electrified, whether for public or private transport. This will represent a significant shift in the demand and supply sides of UK electricity generation.....

(i've not read the book but) I do wonder if the models and predictions about the country's future transport needs are valid.

How we might organise society in future and the choices we might make, including the country's future transport needs, are beyond the scope of the book. However, you might draw some of your own conclusions based on what you read. In the book, MacKay considers the energy consumption of a wide range of transport modes, and the technologies upon which they currently depend and potentially could use in future. The assumptions used are clearly expressed and references are given.

The key point is that we need to quantify what we do now if we are ever going to be able to make sensible and supportable decisions in future with regards to viability, whether it's technical feasibility, affordability, practicality, or any wider consequences for society.

If you want to look our potential options for the future, you might want to look here (weblinks below) from the now-defunct Department of Energy and Climate Change (DECC). I believe David MacKay can be credited with much of what exists behind this Pathways to 2050 modelling tool from when he was Chief Scientific Adviser to DECC. I've not used the new tool (I didn't know it existed until looking for the old tool just now), and it's a long time since I last used the old modelling tool.

http://2050-calculator-tool.decc.gov.uk/#/calculator

This is the original link to old Pathways to 2050 modeller:
http://old-interface.2050.org.uk/pathwa ... ergy_chart

meic wrote:So it isnt actually that relevant to a conversation about the UK's electricity consumption and generation capacity.

Have a look at the book, and you'll see it is very relevant to this debate - probably more than half the book is devoted the UK renewable electricity generation potential. Also, for example, some of the foregoing discussion in this thread has been about transport, which will increasingly be electrified, whether for public or private transport. This will represent a significant shift in the demand and supply sides of UK electricity generation. Other aspects of modern life are likely to become more electrical energy intensive, so the demand for electricity might be expected to increase, even if overall energy usage declines.

meic wrote:The easiest way to reduce that figure is to stop buying useless tat, especially imported useless tat.

Again, I'd recommend you read the book. If I remember correctly, this aspect forms part of MacKay's 'Stuff' category, together with a lot of things besides, such as raw materials manufacturing, packaging, transporting stuff, water treatment and transport, house building, road building, etc. It is undoubtedly true that per capita UK energy consumption could be reduced by avoiding buying tat, and that's tat of any type or origin. The question is by how much? MacKay estimates the respective contributions of each of these to the 125kWh of energy consumption per day per person. If you have a reasonable understanding of the numbers, you can get a better idea of what's involved in making any changes to both the demand and supply sides.

However Brian was talking about 125kWh per person per day.

Brian's recommendation to read David MacKay's book, "Sustainable Energy - without the hot air" is a very good recommendation. It is arguably one of the most important books published on the subject and lauded for its accessibility and objectivity. It is available as a free download, but I strongly suggest buying the physical book - my first edition is now very well thumbed. I have no connection with its publication, and my only vested interest is for promoting a well informed debate in the society, of which I am part.

MacKay explains very clearly the breakdown of the 125kWh figure - it covers all energy use, including for imported goods, and not just domestic electricity consumption. MacKay's approach is to consider each element in turn building up the per capita UK energy consumption. In parallel, he builds up the maximum renewable energy potential of the UK (from all sources) based on the physics and without regard to engineering and economic considerations i.e. the maximum available energy if we weren't restricted by inevitable engineering inefficiencies, the realities of having the to pay for it, or the practicalities or the consequences of delivering it. MacKay leaves the reader to decide what to think about the challenges we all face.

Here is Brian's link again:

The sadly recently deceased David MacKay wrote a great book setting out the challenges of a truly renewable energy supply for the UK.

It is a quantitative (with numbers) analysis, rather than the qualitative “a lot of wind turbines” that many would use.

Many “green” ideas like wave power, hydro and solar simply have insufficient of the necessary resource to make a significant difference. Resource in this case being sea area, rainfall volume/altitude and land area.

Essentially our options come down to build lots of nuclear, industrialise the countryside, or continue to fry the planet whilst running out of fossil fuels. Take your choice.

Some of the numbers are a bit out of date, but it gives the general picture very well.

https://www.withouthotair.com/

I was very saddened to hear of David MacKay's death in April this year.

I've just noticed your latest message, Dave.

I'll try to answer to your questions.

We are principally touring cyclists, and prefer to travel with full camping kit in wildness and mountainous areas on unmade roads. Our fleet comprises steel-framed bikes with mudguards, front and rear racks, and variously equipped with Carradice saddle bags, Ortlieb panniers, and small and large barbags as required. So, we are not particularly fast cyclists and, even when we are 'pressing on', we tend towards the 'full value' or touring end of audax riding too. We like the freedom to ride into the night, and modern hub generator-driven LED lighting is truly wonderful, and provides for much of our USB charging needs too.

I've not seen the evidence that the Reactor requires a higher speed than the Plug for a given output; if anything I thought I recalled reading the contrary. Just looking now, Sinewave claim a peak USB charging current of 1A at 14.4kph whereas the Cinq 5 claim the Plug III only reaches 5W output (which would be 1A at a USB nominal 5V) at a little more than 30kph. Certainly, my experience is that the Reactor performs better than the Plug at touring speeds.

We use various Schmidt SON hub generators i.e. SON 28 Klassik, new style SON 28 and the SONdelux, and even one of the original (first generation) SON hubs, which must now be over 15 years old. For slower touring speeds, a SON 28 is our preference. Our standard set up is to connect them to Schmidt Edelux (an Edelux II in one case) front lights with B&M Secula or Toplight Line Plus rear lights. The lights are controlled automatically by the Edelux ambient light sensor. As I mentioned, we keep our Garmins connected to our Sinewave Reactors all the time, and sometimes via a cache battery. The cache battery can also be used to charge phones, cameras, headtorches, etc. when touring too, and can be topped up from the mains when the opportunities arise. It all works very reliably throughout the year whatever the weather.

As an aside, and since someone mentioned it earlier in this forum thread, I have used the B&M Luxos U with USB charging too, and my experience is that the Reactor has a far better USB charging performance than the Luxos U. This difference in performance is greatest at night, when the Luxos wouldn't provide any USB charging with the light on dipped beam, whereas the Reactor will charge a USB device and power an Edelux perfectly adequately for all but the fastest, winding, unlit descents. If ever I needed more light output from my Edelux, unplugging the USB charging cable would provide a little extra lighting using the Edelux at its full potential. The Luxos is an excellent light, but I find the U-variant is disappointing with respect to its USB charging performance.

Just looking at a Plug, it is clearly intended for 1⅛" steerers. However, I suppose with a bit of thought you could fit one to a bike with a 1" steerer, but you would need to have a 10-15mm column of 1⅛" headset spacers on top of your stem to provide access and clearance for the wire and connector from the generator (dynamo) to slot into the underside of the Plug. It's not ideal though.

However, there are several reasons why my preference has been to choose the Sinewave Reactor over the Plug, and the arrangement I describe above would exacerbate one of the design weaknesses of the Plug. The Reactor puts almost all of its electronic gubbins inside the steerer and adds only 9mm to the height above the stem, whereas the Plug sits on top of the stem and adds 25mm to the height plus the 10-15mm of spacers that you'd need. Whether this is a problem for you will depend on whether you mount anything on your stem (e.g. GPS, phone, etc.), the profile of any GPS/phone mount and the length and angle of your stem. Unless you can mount a device well forward of the Plug, the device could interfere with the Plug and so effectively rule out using the stem for mounting such devices.

Furthermore, and more importantly, my experience has been that the low-profile Reactor has proved far more reliable than the Plug. I have 3 Plugs (all bought secondhand at different times, I should add, so I don't know their full histories) and none of them works anymore; the problems are all with the non-serviceable 'head unit' rather than the replaceable PAT cable. Meanwhile, our 5 Reactors (all bought new and at various times since the product was launched in late Spring 2014) continue to perform fauItlessly.

I am doubtful of the the Plug's weatherproofing. As an example, last December (2014), my partner and I were cycling in a very wet Sri Lanka for three weeks, and I mean very wet as in nearly 24 hours-a-day heavy tropical rain for 10 days. If and when it ever did stop raining, then we were usually off the bikes in our various hotels asleep! We had no option but to move on and ride each day and we generally managed to keep ahead of floods, evacuations, landslides and road closures; it was pretty miserable until we reached the south coast and the sun. I was using a Reactor and my partner was using a Plug III. Contrary to common advice, and for 7-8 years now, we've generally ridden with our Garmins plugged into various USB chargers all day and every day whatever the weather and, despite taking steps to weatherproof our setups, these extreme Sri Lankan conditions meant nothing remained totally dry. By the end of the holiday, the Reactor, the Garmins and USB cables were still all working perfectly fine, but the Plug had stopped working.

We'll never know if the damp caused the Plug III's demise, but a third failure out of three units is informative. Perhaps I have been unlucky with the Plug, but draw your own conclusions - I have certainly drawn mine.

We have continued to use Reactors throughout a wet 2015 PBP qualification season in Scotland and the north of England without any issue keeping our Garmins fully charged day and night; the Reactors have been entirely trustworthy. Thankfully, PBP was dry!

Here's an update on this bike.

This bike was stolen during a house break-in in January 2014. It had been missing for ~1.5 years when I noticed it was listed on eBay. Someone also brought it to the attention of any interested forum members in this 'For Sale' thread (viewtopic.php?f=40&t=99067), but no-one had noticed it was the bike I had reported stolen in this thread.

My friend, the original owner, contacted her local police, and they were very interested to recover it. They recommended not alerting the eBay seller by any direct contact or by posting anything on this forum. My friend also contacted eBay, and the police were interested to obtain the address of the seller from eBay in order to recover the bike. This proved to be a slow process and, when it became clear eBay would not have responded to the police by the time the auction closed, we decided to warn CTC forum members (which was condoned by the police), and I posted my warning in the 'For Sale' thread.

As a result of this warning, a CTC forum member contacted the eBay seller. The seller was apparently unaware it was a stolen bike, and said he would take the bike to his local police station when he returned from holiday. All credit to the seller, since he did exactly that.

So, the bike was recovered. However, since it was more than a year after the theft, my friend had already bought a replacement bike, and the actual owner of the stolen bike was now my friend's insurer. The insurer was also keen to recover the bike, so that it could realise some of the value. Therefore, I expect the bike will be sold by the insurer over the coming months, but I don't know where or when.

Dave W wrote:

andrewjoseph wrote:auto front shift was known about at least 14 months ago.

http://road.cc/content/news/119890-shimano-introduce-synchro-shift-video

Now that, would be superb on a tandem.

It is! And, that's using a 44-34-24 triple and not the the standard Shimano MTB Di2 40-30-22 set up.

Brucey wrote: the rider would need to know if the system was 'thinking of a double shift or not' beforehand I reckon.

Yes, you get a short 'beep' before a double shift. The shifts are crisp anyway, so it doesn't seem to matter if you forget the next shift is a double shift whether going up or down.

Brucey wrote:autoshift would presumably need to be programmed for different gear setups, and I wonder what happens if you ask for a 'block shift' or multiple shift of some kind? -hopefully the mechs don't make many needless movements?

Yes, you can use the two standard Shimano sequential shift programmes, or design your own, which is what I've done. You can also shift the front mech with the left-hand switches independent of the 'syncroshift' programmed shifting if you want to anticipate a steep hill, for example.

Colin Stanley wrote:Well, can anyone explain how it would work with a 10 speed Shimano rear mech and Campag' Ergo levers?
This is planned to be the setup for my new tandem using a ShiftMate in the cable run. The frame will have couplings to disassemble it for holiday journeys, so the front and rear gearshift Bowden cables will have joining connectors (a bit like an electrical plug and socket?).
I guess electric shifting is out of the question for such a setup unless it would be possible to match the signals from both products?

Brucey wrote:Incidentally there is a positive reason for wanting to use Di2 on a tandem and that is that it might make the shifts more consistent ( long Bowden cable runs are often more troublesome) and separation on your machine might be easier if the right connectors can be used

It depends what your reasons are for a Shimergo set-up. Usually, the choice of Shimergo is to dispense with the indexed Shimano front shifter, and use the Ergo's micro-indexing to allow better trimming of the front mech. However, Di2 trims automatically as you move up and down the cassette. So, if that's the reason for choosing a Shimergo, then that requirement is obviated with Di2. I previously had 8, 9, and 10 speed Shimergo set-ups on the tandem and all worked very well. However, as any tandem pilot knows, a typically hilly route requires frequent gear changes, and after 20-30 or more hours' nearly continuously riding on long brevets, one's hands become tired and weak from all the shifting. Di2 is a real boon, and shifting is more reliable than ever, although that wasn't my reason for the conversion to Di2. We also use a coupled tandem (our frame splits into three parts), and this is almost trivial for the Di2's CAN bus system - all you need is an extra junction box and extra cable to connect the two junction boxes.

Regarding compatibility, I believe the road and MTB systems are mutually compatible, except you can't mix front and rear mechs. We use road shifters and MTB mechs; the rest of the system is common to both. I know some tandem riders would like a road front mech suitable from bigger chain rings and an MTB rear mech, but I gather this doesn't work at all. Of course, there isn't a triple road front mech, so the MTB is currently the only way to go if you want a triple system. I'm aware that someone has re-modelled the cage of an MTB front mech to make it work with a typical size large road chain ring. Also, the 11-speed MTB cassette doesn't need an 11-speed specific hub - the largest sprocket is so large the rear mech is well clear of the spokes.

There were lots of comments about charging in this thread too. Well, it's not much difference from charging your phone or Garmin, except you need to do it far less often. Any USB power source is sufficient, whether a hub-generator powered devices (such as a Cinq5 Plug, Sinewave Reactor, usb-werk, or whatever), a cache battery, or mains-powered charger. It's quick and simple. Our experience riding hilly undulating roads requiring lots of shifting is that one can probably readily ride ~1000km between charging, not that we've tried. Apparently, below 5% battery level, the system switches to 'limp' mode when front shifting is limited to changing down to the middle ring and rear shifting works as normal until at some point you're restricted the gear you're in when the system stops shifting all together. There's plenty of warning with a battery indicator on the handlebar display.

There's no denying Di2, especially in the MTB form, is expensive (but it is already being discounted 25-30% or more). We'll have to see how long the battery remains serviceable, but my experience is quality Lithium battery systems have a long service life if they don't get too hot - I still have an 8-year-old Garmin with an internal battery that can still manage 12-15 hour days. Of course, there's always a chance of a duff battery failing prematurely. However, the cost of a battery is not much different from a pair of decent tyres. Like other commenters in this thread, I'm also somewhat sceptical about 11-speed chain life, especially on a tandem - we'll just have to see. For belt and braces when it's important to have a fallback, I've rigged up a non-indexed down-tube Bowden-cable shifter for the rear so that I can use a normal rear mech should the need ever arise. Of course, I'd still have to find a cable-operated MTB rear mech in that event too, but it's some insurance against a complete Di2 failure.

Di2 is not for everyone to which as this thread is a testament, but I thought it may interest some readers to hear a contrary view based on hands-on experience.