Prius tank-by-tank economy report

When I calculate fuel use as I fill the tank I’ve been averaging 4.9l/100km (57.65 MPG(UK) 48.23MPG(US) 20.41km/l with HSI indicating 4.6-4.7l/100km).  This is not as good as I had hoped.  But then I replaced the dead 12 volt battery and pumped the tyres up to 42/40psi from the low 30’s that they had been at.  Expecting far better fuel economy I was disappointed to return 4.8l/100km on my last tank (HSI indicated 4.5l/100km).  A few ideas come to mind:

  • The Prius fuel tank is actually a bag designed to prevent fumes and thereby reduce emissions.  It seems to vary in size every time I fill it.  Despite being a 45 litre tank, the most I have been able to put in is 35.21 litres and that was with a very slow fill.  Maybe I’m not filling it to the same level each time?
  • I let my brother-in-law drive for about 40km, at city and highway speeds, with bursts of full acceleration and braking.  I should have set the trip meter just to see what he used.  A rough calculation suggest that if his trip was at 8l/100km, then that would change is 4.6l/100km tank into 4.8l/100km.
  • The last tank was Shell E10 94RON.  If I’m generous and only allow 3% reduction in economy, then that explains 0.15l/100km of the difference.  (Current tank is BP 95 RON)

The ADR 81/02 test combined figure is 3.9l/100km.

What’s the ADR 81/02 test?

The ADR 81/02 test is run on a dynamometer with account made for the aerodynamic characteristics and weight of the car.  The city cycle runs for 13 minutes and is a stop-start ride with speeds up to 50km/h.  The car spends a total of four minutes stopped during the 13-minute cycle.

The extra-urban cycle involves the car accelerating from a standstill and holding various speeds up to 120km/h.  for some reason the car is stopped for 40s of the 6 minute 40 seconds extra-urban test cycle. has more details, but you need to know your maths and stats to make sense of it all.  I’m still trying to determine if a Plug-in Hybrid is tested using EV mode or if it is forced somehow to behave like a standard hybrid.  Read this and tell me, if you could.

Does the ADR 81/02 test have any basis in reality?

Most fuel economy tests employ conditions that could be considered unrealistic.  For a start, when you drive somewhere, you don’t stay in a shed on a rolling road at a constant temperature; you move around going up and down hills, with and against the wind and on varying road surfaces.  However, the tests are reliable in the sense of being repeatable and comparable across a wide range of vehicles.

Now, on this planet…

In the real world, there are many more variables.  It is difficult to compare the consumption of one tank of petrol to the next unless every day is the same journey in damn near identical conditions.  Driving style can account for 30-40% of fuel economy, usually for the worse.  Very cold temperatures can affect economy as it takes longer to warm the engine and fuels sometimes have additives that keep the fuel stable but not as potent.

The lowest figures for the Prius are around 30 MPG(US) 7.88l/100km 25.1 MPG(UK) 12.74km/l in extreme circumstances such as sub-zero temperatures and short journeys or holiday trips with a full load and roof racks.  There was a journalist who flogged a Prius (previous NVW20 model with the 1.5 litre engine) and got about 30MPG(US) despite driving like maniac.

Next steps?

I have to work on my Pulse and Glide technique.  I already found that my pulsing is not as high and my gliding is not as far as I could do it, traffic permitting.

I also realised that a fast uphill trip on Majura road used a lot more fuel than a fast downhill trip with a return via the city.  Maintaining 90km/h on a rough road while climbing 130m and descending 60m might not be as fuel efficient as travelling 60-80km/h with more gradual climbs; traffic lights notwithstanding.

I’m tempted to try some more extreme measures if this tank of BP 95 RON doesn’t perform as I hope.

Prius – what is it like on the highway part 1

First long trip

Today was the first long journey in the Prius (still unnamed) taking the dogs to Sydney for their grooming and some shopping for us. The Prius was known for having worse fuel economy on the highway compared to the city, the opposite of every other car I’ve known. The difference was slight but it highlighted that stop-start traffic and low speeds suited a hybrid whereas highway driving at 100-110km/h (62-68mph) required the engine to run continually. A modern turbo diesel would be expected to drink less as its torque helps it keep a steady speed regardless of load.

The official figures according to ADR81/02 for the Toyota Prius Hybrid 1.8L NVW30 are:

  • Urban 3.9 l/100km 72.4MPG(UK) 60.6MPG(US) 25.6km/l
  • Extra-urban 3.7 l/100km 76.4MPG(UK) 63.9MPG(US) 27.0km/l
  • Combined 3.9 l/100km 72.4MPG(UK) 60.6MPG(US) 25.6km/l

So that pattern is more conventional.

The route

Gungahlin to Sydney is about 280km (174 miles). The altitude starts at about 620m above sea level (ASL) rises to 695m near the ACT-NSW border, peaks at 756m a few times and remains above 600m for about 165km of the journey.

There's some hills between Canberra and Sydney

Route profile from Canberra to Sydney, courtesy of Google Earth

Until I had seen this profile I had no idea how gradual the slope was. Not very. Interesting also that one climbs 1553m and drops 2144m in the process. Now, the profile is very important to explaining the fuel economy figures, which I’ll get to in a minute.

The measured trip – out lap

The trip started from filling up with 31.6l of 95RON at BP Marulan, about 114km (71 miles) from home and 638m above sea level. From there to Sydney has 810m of climbs at a maximum of 4.6% and 1418m of descents at a maximum of 5.2%. So that stretch happens to include the steepest climb and descent of the entire Canberra to Sydney journey.

I used the cruise control at 110km/h by GPS (114km/h on the speedo). PWR mode was used to enter from ramps but I used ECO mode for the rest of the time. I tried to anticipate steep hills by accelerating before, but I wasn’t a good judge of momentum so the cruise control kicked in to full power mode to restore speed… not good for economy. I didn’t draft or use hypermiling techniques, but I was smooth. Top speed was a burst at about 120km/h (the de facto speed limit on the Hume Hwy) to make way for a another car. The A/C was low at 21.5°c (71F) against 15°c (59F) outside.

I found that on many descents, I was rolling and recharging the battery using zero fuel and maintaining cruise control speed. (Note: the engine must run at those speeds to avoid a large discrepancy between the road speed and transmission speed. It just does it without using fuel, just as a conventional engine might.) With plenty of chances to recharge, there was plenty of battery for hybrid eco gliding.

The last part of the out lap was an EV drive through the underground car park at Cathedral Street Sydney. (On the return journey I nearly used EV for the whole climb back out of the car park, but the engine started with about 50m to go.)

The trip computer read 160.5km (99.7 miles) 4.0l/100km 70.6MPG(UK) 59.1MPG(US) 25km/l with average speed 92km/h (57.1mph). I was very happy and a bit surprised. (That’s not the 168km from Google Earth as I took a more direct route through the suburbs to drop off the dogs and then to the bookshop.)

The measured trip – return

Conditions were a bit warmer for the return, so the A/C was set to 23°c (73F) against 28°c (82F) outside.

At the M5 entrance at Arncliffe the traffic was banked up. It took 10 minutes to drive (in EV) the 250m from the Marsh Street entrance to finally get onto the M5 itself. Soon after joining the stream the traffic sped to 60km/h! Bad merge genes?

Again I stuck to 110km/h (68mph) for the run back. We took one stop at BP Marulan for the dogs to takes a break and get some Hungry Jacks (Burger King).

Note: I might have been best to fill again at this point and get a reasonable downhill/uphill comparison, but I wanted to get home.

But wait! Oddly enough the elevation profile on the return is different. The climbs total 2064m (c.f. 2144m 4% more) the descents total 1465m (c.f. 1553m 6% more) and the maximum slopes are both 3.9% (c.f. 4.6% climb 5.2% descent). The north- and south-bound lanes are divided and in some cases separated by as much as 300m. Perhaps the grade was engineered to be less severe on the “uphill” journey?Going uphill is easier than going downhill?

Going uphill is easier than going downhill? Not quite

Arriving home the trip totals were 450.1km (279.5 miles) 4.6l/100km 61.4 MPG(UK) 51.4MPG(US) 21.7km/l Ave. 78km/h (48.5mph).

The smoother ride home helps explain why I rarely noticed recharging while descending; the slopes weren’t steep enough to maintain 110km/h without the engine.

Interesting preliminary results. We return to Sydney next week for a few days, so I’ll get another go soon. I leave you with this image as I stopped in my driveway with a Ford F250 truck in the background.