The 2017 Power Unit usage review

Photo by:

This has been the fourth year of the current Power Unit (PU) regulations, and although it has seen the performance closing between the manufacturers (even Honda were looking competitive with the midfield at the end of the year), there still seems to be a significant difference in terms of reliability.  As the number of Power Unit elements allowed is reducing in 2018, this may become a deciding factor in the outcome of both world championships next season.

As many of you are aware, the Power Unit is comprised of six elements:

  • Internal Combustion Engine – ICE
  • Turbocharger – TC
  • Motor Generator Unit – Heat – MGU-H
  • Motor Generator Unit – Kinetic – MGU-K
  • Energy Store – ES
  • Control Electronics – CE

In 2017, each driver was allowed to use four complete Power Units during the season before incurring any grid penalties.  The elements could be mixed and matched as the season wore on to enable teams to gain the maximum life out of each component before receiving penalties.  Sometimes however, the time required to change just one element of the PU exceeded that available (particularly between Free Practice 3 and Qualifying) and it was quicker to swap out the complete ICE, TC, MGU-H, MGU-K assembly.  This action may have incurred penalties earlier in the season than absolutely necessary.

The FIA reports available just state the total number of each element used by each driver/car at each point in the season (the start of a race weekend).  They do not state which particular element was used in any particular session (except for the very first time that element is used).  It is therefore impossible to get accurate data of the life of each component, and the best that can be done is to calculate the average life of the components used on a particular car.

Note that although I have referred above to the driver having an allocation of four PU for the season, it is really the car that has this allocation.  Carlos Sainz started the season at Toro Rosso and finished it at Renault.  Although he used Renault Power Units throughout the year, these did not transfer between the teams, when he went to Renault he took over the allocation that had been used by Jolyon Palmer to that point.

For 2018 each driver will get three ICE, TC and MGU-H for the season and just two MGU-K, CE and ES.  This is a substantial reduction on the allowance for 2017, and will require another increase in reliability if the outcome of the title is not to be impacted by grid penalties.

In the following graphs, each coloured bar represents a different race in the 2 race season, and the height of the bar represents the total number of that element that the driver has used until that race.  The driver’s names are those as at the end of the season.

Internal Combustion Engine – ICE

Honda obviously gave up trying to stay within the regulation limit, and instead concentrated on the development of the whole PU to deliver a more competitive power output.

The Renault powered cars all exceeded the allowed four ICE for the year, and although at times they appeared to be competitive (winning three races during the year in the back of the Red Bulls), this may be due to the aero efficiency of the car rather than the power output.

Both Mercedes and Ferrari had one car (in both cases the one leading the title chase) that exceeded the maximum allowed number of ICE.  In both cases it was a tactical change, when the driver was already starting from the back of the grid due to problems in qualifying.  In the case of the Mercedes, this fresh PU certainly seemed to overcome the difficulty that the car had elsewhere when following others.  The step in performance wasn’t so apparent  in Ferrari’s case, but it did occur earlier in the season, and so the PU would have been expected to do a higher mileage than the late replacement Mercedes unit.

Turbocharger – TC

The story is very similar for this element.  Honda used even more TC than they did ICE, every Renault driver exceeded the allocation and only the leading drivers from Mercedes and Ferrari exceeded their allocation, again tactically.  However given how early in the season Vettel took his additional ICE, TC and MGU-H he may have needed to take these additional elements anyway.

Motor Generator Unit – Heat – MGU-H

This was Renault’s real weak point, and particularly Toro Rosso suffered with these components.  They used three times as many as they will have in 2018 if they are to get through the season without penalties.  It wasn’t just Toro Rosso though, as Daniel Ricciardo needed eight to last the season.  The other manufacturers were the same as the TC usage.

Motor Generator Unit – Kinetic – MGU-K

Only Honda exceeded the four allowed elements for the season.  Renault in stark contrast to the other MGU only needed three of these components for all bar one of their cars.  It is hard to know if going down to just two of these will resent real problems for any of the manufacturers (this MGU is linked to the crankshaft, so spins significantly slower than the MGU attached to the TC).

Energy Store – ES

This has historically been one of the most reliable of the PU elements, along with the CE. While Honda have no doubt been changing the specification of their complete PU through the year in order to try and catch up on the overall power output, the fifth ES required by Brendon Hartley is worrying for Renault next year.

Control Electronics – CE

A similar message for Honda and Renault as for the ES, although both Toro Rosso’s needed a fifth ES, so perhaps the problem is slightly worse?  Mercedes on the other hand look in good shape for 2018 with this element, as half their drivers got through 2017 on just two ES, the limit for 2018.

The graphs above show how many of each PU element each car used through the year, however to estimate the reliable life of each component we also need to know how far the car drove during the season.  The table below shows the distance covered for each car during the race weekends (so the periods when the number of PU elements were restricted by the regulations).  It includes the distance (kilometres) covered by third drivers in free practice and any replacement drivers but doesn’t include any of the pre-season, in-season or post-season tests.  As before, the cars are identified by the race drivers that finished the season driving them.

Just on the distance covered, it seems straightforward to rank the PU manufacturers in order of reliability.  On a trouble-free weekend, each car would cover over 700 km, the races (excluding Monaco) are approximately 310 km each, the rest of the distance is free practice and qualifying.  Evidently a driver who is eliminated in Q1 is likely to do fewer laps than one who gets through to Q3, but those at the back of the field often have three attempts to set a lap time in Q1, while those expecting to fight for pole will only need one run.  The main difference in distance run is then due to either mechanical problems or accidents on track.  The total number of race laps each driver completed is shown in this table:

All the Mercedes powered teams completed a good distance, although it looks like Williams may have been limiting their mileage when compared to Force India.  However they did complete fewer race laps than the Force India team, so it looks like they retired more often (or at least earlier in the races) and this may explain the difference.  It is worth noting that while Lewis Hamilton was the only driver to complete every race lap this season, he was over 300 km short of the total distance recorded by Esteban Ocon.  This is likely due to Ocon’s inexperience meaning he wanted to do more laps in practice.

The Ferrari powered teams had some accidents / incidents in the races which may go some way to explaining the shorter distance they completed when compared to the Mercedes teams.  Only the works team on a par with the Williams outfit.  The anomaly with the Ferrari powered cars is the comparison between Haas and Sauber.  Haas drivers completed far more race laps than the Sauber drivers, yet a shorter distance over the entire season.  With neither team exceeding the permitted four PU per car, it looks like the year old PU in the Sauber was more reliable (if slower), while the Haas team limited the mileage in Free Practice to save the PU for the race.

When it comes to the Renault powered cars the figures are reduced due to the unreliability which caused so many grid penalties to be applied.  It is worth noting that while the race laps table is for the driver, the total distance table is for the car.  So Carlos Sainz – who drove two cars through the season appears to have a low total distance as he ended the season at Renault driving the car that Jolyon Palmer had used the majority of the year.

The Honda powered McLaren’s are much lower in the total distance table than they are in the race laps completed table, indicating that they either had significant problems in Free Practice, or they were saving mileage for the race.

As stated above, the FIA reports produced for PU use only show the total number of elements used by each driver at each race of the season, they do not state which component was used in each session.  To estimate the component life I have taken the total distance covered by each car and divided it by the total number of each element of the PU used over the course of the season.  With the top three PU manufacturers this gives a reasonable spread of data with which to estimate the likely life of each element.  With Honda, the data is over a much smaller sample (two cars rather than six) and so is likely to be more prone to error.  The following graphs show the calculated life of each element of the PU, with bars showing the range of values for the different cars using that PU.

All the graphs have the same vertical axis, so it easier to compare the ranges of each manufacturer’s PU.  With the drivers completing over 15,000 km this season and an extra race next year, each ICE, TC and MGU-H will have to last 5,300 km and each MGU-K, ES and CE will have to last 8,000 km.  As can be seen from the graphs above even for Mercedes this is further than they have gone so far, but for Ferrari and Renault it could be a much bigger problem.  The Honda figures may be affected by the decision to keep developing the PU to try and get a competitive power output rather than trying to minimise grid penalties.

Lewis Hamilton’s performance in Brazil, with a fresh PU only required to last for the race in Brazil and the weekend in Abu Dhabi (a total of less than 1,200 km) shows just how much the current generation of PU are restricted in order to provide the desired reliability.  Ferrari went through PU components at a faster rate than Mercedes at the start of this season, yet managed to finish the year using the same number of components as Mercedes.  However, did the teams competitiveness drop later in the year as they made the limited number of components last until the end of the season?  Did Red Bull’s increased competitiveness at the end of the year benefit from relatively late additional PU components being fitted to both cars (which needed to last for a much shorter distance)?

In 2017 we have seen the power outputs of these PU converging, with Mercedes no longer appearing to have the power advantage it did when these regulations first came in.  This is encouraging, but the title chase was impacted by driving/operational errors from Vettel and Ferrari and so we were saved from the potential impact of grid penalties deciding the championships.  There is work for all PU manufacturers to do over the winter if grid penalties are not going to decide the outcome of the 2018 title.


0 0 votes
Article Rating
Notify of

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Newest Most Voted
Inline Feedbacks
View all comments

Salvu Borg

The work needed to be done by the PU manufacturers for next season longer duration PU elements would have been already signed off by this time of the year.


Excellent information and analysis. Thank you.


“Lewis Hamilton’s performance in Brazil, with a fresh PU only required to last for the race in Brazil and the weekend in Abu Dhabi (a total of less than 1,200 km) shows just how much the current generation of PU are restricted in order to provide the desired reliability.” Well said. I would expect that the apparent ceiling at the ~3800 and ~3600 km level for at least the first four components for Mercedes and Ferrari was a product of your statement quoted above and not an absolute upper bound. Even so, your conclusion remains undoubtedly true. Though I would… Read more »


Very cool article and analysis. I’m curious what data you are using to draw this conclusion “In 2017 we have seen the power outputs of these PU converging, with Mercedes no longer appearing to have the power advantage it did when these regulations first came in.” In every single track that had a long(er) straight or set of straights, Merecedes beat the Ferraris. In addition, Renault and Honda at those tracks were even farther behind. I agree that there is a slight convergence, but to me Mercedes still have the clear power advantage in addition to their reliability advantage. Although… Read more »


Despite Hamilton being usually faster in qualifying, Vettel managed to lead the drivers title for the first twelve rounds of the series.
Certainly in the first half of the season there wasn’t much between the two teams in terms of overall performance. This is a step change from recent season’s when the Mercedes were clearly in front.
Then at the end of the year, Red Bull won two races on merit.


Agreed, VET did manage to lead the championship through the first half, but I would argue that was due to 1) HAM’s relatively poor performance in races that he didn’t win – Specifically Russia (4th), Monaco (7th), Baku (5th, but not his fault), Austria (4th), Hungary (4th) and 2) Ferrari having (for whatever reason) a better car on tighter/shorter tracks. I think it’s very hard to draw the conclusion based on these facts that Ferrari had the same power as Mercedes. To add to it, the 2nd half of the season really proved that with all of the power tracks… Read more »


For the first part of the season Hamilton certainly looked like he had a couple of off weekends (which cost him the championship in 2016 just as much as the reliability issues). However his driving for the second half of the year was much improved, and for the first time that I can remember he didn’t resort to the ‘Schumacher/Senna’ approach of driving their rivals off the road. Ferrari and Vettel in particular looked the better combination in the first half of the year, but to do so they had to use PU elements at a faster rate. To try… Read more »

Salvu Borg

Dave, Prosit for your graphs and bars, very neat indeed. Just in case you haven’t seen them, and just in case you are interested, lately there were some incredible Honda power unit photos, incredibly clear and showing fine details, the likes that has never been seen before, the latest set of 12 photos originated on a Czech F1 site (F1news.autoroad.CZ.) These first set were not very clear, but somebody on another English speaking F1 site not only managed to extract/lift them, but made them incredibly clear as to show the fine detail, a really incredibly compact and complicated build engine.… Read more »

Salvu Borg

Totally off-topic but might interest some on this here community, @ NC.
Some time ago the relevance of the MGU-H part of the present F1 hybrid to road cars use was questioned, at that time I said that BOSCH are spending huge sums of money on such research programs.
Now last week (Dec 7. 2017) Honda filled for Patent for a road car MGU-H.
Dec 7. 2017 United States patent application publication kamio. Pub number US 2017/0350315A1. Pub date Dec 7. 2017.…#view=fith.

Negative Camber

That’s great, I hope they discover an affordable way to add a MGU-H system to a road car that isn’t subsidized by tax payer money and the public wants to buy based on merit and performance. As with technology evolution, going backwards in performance isn’t the trajectory arc of many evolving technologies…I’m not sure why the imposition is being forced upon the public. When hybrids, electrics and others effectively directly replace or even out-performa current options, the public will always be reticent and pouring tax money into it isn’t the best idea in my opinion. Still, very cool they are… Read more »

Salvu Borg

Yes that’s great, any effort to improve the road car efficiency should be lauded. This efficiency improvement even greater because it is being gained by means of energy that is otherwise going to waste.
The formula one manufacturer’s efforts to develop and improve the present power unit efficiency is certainly drippling down to road car use.