Read here: http://www.lamptech.co.uk/Documents/FL%20Gases.htm .
The Eco or energy savers T8, T5 and PL lamps probably have pure xenon buffer, instead of the argon-krypton penning.
Unfortunately, the energy saver fluorescent lamps, are very sensitive to ambient temperature variations, compared to argon and krypton buffered fluorescent lamps.

So except electrode wear they are good as XXXXXXVHO xenon strobe tubes ?

Read here: http://www.lamptech.co.uk/Documents/FL%20Gases.htm .
The Eco or energy savers T8, T5 and PL lamps probably have pure xenon buffer, instead of the argon-krypton penning.

The article say nothing about pure Xenon and I doubt pure Xenon would be feasible.
The main reason would be the too high striking voltage.
I would rather guess it is again some Penning mixture, that time involving Xenon as one of it's components.

The main use of these I would not see their higher efficacy then standard tri-phosphor T8, but taking the lumen output back towards the "original" halophosphate T12 40W, what many (if not most) installations are designed for, so save the energy use by avoiding overillumination.

But the same effect could be done in much cheaper way using standard T8 tri-phosphor tubes, but replace the 0.43A "36W" ballast by the 0.36A "18W" one. That setup is even more efficient, as "18W" ballast have way lower losses then the "36W" one, moreover when the "36W" one is overdriven (but only slightly) by the lower arc voltage lamp.

So the Eco T8s, T5s and PLs probably contains argon-krypton-xenon penning.
Edit: I already thought that the primery reason for the Eco fluorescent lamps, is to take the adventage of the high efficiency of triphosphors for less wattage instead of more light, in order to prevent overlighting, as the tripohsphors 36W T8 have higher lumens than the halophosphors 36W T8 and don't yield any energy saving.
In this way of the ECO fluorescent lamps, the conversion to triphosphors will save energy over halophosphors, and still get the higher lighting quality of triphosphors over halophosphors.
The american equivalent lamps (F34T12), don't uses xenon in their buffer (Just argon-krypton) and still have halo/deluxe phosphors and are rapidstart lamps (They have the electrically conductive coating of the F32T8), so not surprisingly that they don't yields any energy saving while distroys their F40T12 ballasts of any type.
The european version of the lamp, don't expected to do this.

If it will have lower voltage, it will take higher current on all series inductor magnetic ballasts - Switch Start and Perfekt Start, and cause the ballasts to overheat. It would also severely damage Perfekt Start ballasts by overpowering the heater in the thermal relay

If it will have higher voltage (to the extent that its more than 1/2 of the mains) it would save power and make the ballasts run cooler (and save power in ballast losses too) but i am unsure whether it would even work with higher arc voltage

What they could do is, require an additional magnetic ballast to go in series with the existing one to drop the current and then the lamp would do well with thew existing Ar-Kr filling without need for Xe anyway

They do have lower arc voltage and do save power in that way.
The current increase by using 32W instead of 36W lamp is about 2..3% (include core nonlinearity; way less then the mains tolerance causing more then +/-10%), so it could not be the reason for ballast to overheat. So if the ballast does overheat, there is for sure some other wrong thing to look for.

I hope that these Eco fluorescent lamps will be available in Israel, as if i will use a single T8, and a single 18W T8 halophosphors will produces enough light in my room, so i will switch it to T8 Eco 16W, when it will reach EOL, in order to retain the same level of brightness, but to reduce the power consumption and still be benefited from a high quality white light of triphosphors.
However, if the single T8 halophosphors, will be too dim, i will had to buy a regular triphosphors 18W lamp.
Because the T8 Eco, are only saves energy, when used as a retrofit of halophosphors T8s (Similar intensity as the halophosphors T8), i don't know why there are also T5 HO/HE Eco and PL Eco available.
Btw, Philips lies when they states that their MASTER TL-D Eco lamps offers the same light output as their MASTER TL-D Super 80 lamps, instead of their regular halophosphor TL-Ds.

@dor123: In your place I would use some lower current ballast and standard 18W tri-phosphor tubes.

These "ECO" would be terribly expensive and mainly with single 18W you won't have any noticeable difference in power input, as the ballast dissipate nearly about the same as the lamp...
If your fixture is 2x18W, use "1x18W" ballast (instead of "36W" one).

For single F18T8 use pair of "PL-S 7..11W" ballasts in parallel (you may switch one OFF after lamp start for dimmed setting; Then use "..80W" starter), that would lead to ~0.33..0.34A, what mean about 16W on an 18W lamp.

I can use 18W T8's on either 2x 1x18W ballasts or 1x 1x36W ballast in series

Will it affect the amount of 100HZ flicker from the lamps ?

My guess is that it might, as 1x18W lamp would stay on for longer part of the AC cycle. But i dont know how significant the effect would be

I can use 18W T8's on either 2x 1x18W ballasts or 1x 1x36W ballast in series

"2x18W" is the same ballast as "1x36W", the higher combined voltage of two 18W in series cause the "36W" ballast to run at about 0.37A.
But if you use single "1x18W" ballast (0.37A with single 18W lamp) with two series connected 18W lamps, the current would drop down to about 0.33A, what mean it would run the 18W tubes at about 16W.

Will it affect the amount of 100HZ flicker from the lamps ?

No at all, only the intensity drop. The flicker would be the same as when run on "1x36W" ballast (assume series pair configuration).

My guess is that it might, as 1x18W lamp would stay on for longer part of the AC cycle. But i dont know how significant the effect would be

On an inductive ballast there would be no effect, as the current shape is nearly independent on the arc voltage, as the current drop to zero when the mains already build up significant voltage ensuring immediate reignition in the opposite current direction. For extreme arc voltages the shape move from sinusoidal (too low) towards triangular (in an "asymptotic" manner at high voltages), while both shapes are already very similar, so even such voltage extremes could not cause any significant difference in the flicker.

Resistive and/or capacitive (without any significant inductor in series) ballasts are way different story, they keep the lamp without any current for quite some time during zero cross, what does cause severe flicker (and lamp wear) and does not allow to operate lamps with too high arc voltage.

I checked Philips website for the operating temperature range of their Master TL-D Eco (Which is an example of a xenon-argon or xenon-krypton-argon buffered ECO T8 fluorescent lamp), and found that this type of lamp is for indoor use only, at an environment temperature of above 20degC (25degC-35degC). Exceeding of this temperatures range, may results in significanly lower light output, and poor lamp performance.

The existing 36/40 W ballasts were designed to run 36/40W lamps, optimized for them, and made with some design margin for stability in wide temperature range, lamp deterioration during its life, line voltage deviation etc

The new lamps are not of the same properties as the old ones, and are meant to fall into the design margins of the ballasts, but not in the same point where the original lamps are

So, some of the margins are way cut down with the retrofit lamps

This is why those lamps have problem in cold temperature (compatibility to existing ballast at the expense of wider temperature range), US 34W lamps kill ballasts (cutting too much into the temperature safety margins) etc

And also the reason why sometimes retrofit lamps can be less efficient than the originals - as they are designed to work on different power ballast, instead of both lamp and ballast being matched and optimized together for efficiency (this holds for retrofit lamp of the same lamp type, not for stuff like new triphosphor vs old halophosphate, or HPS retrofit lamps etc)

The existing 36/40 W ballasts were designed to run 36/40W lamps, optimized for them, and made with some design margin for stability in wide temperature range, lamp deterioration during its life, line voltage deviation etc

The new lamps are not of the same properties as the old ones, and are meant to fall into the design margins of the ballasts, but not in the same point where the original lamps are

So, some of the margins are way cut down with the retrofit lamps

This is why those lamps have problem in cold temperature (compatibility to existing ballast at the expense of wider temperature range), US 34W lamps kill ballasts (cutting too much into the temperature safety margins) etc

And also the reason why sometimes retrofit lamps can be less efficient than the originals - as they are designed to work on different power ballast, instead of both lamp and ballast being matched and optimized together for efficiency (this holds for retrofit lamp of the same lamp type, not for stuff like new triphosphor vs old halophosphate, or HPS retrofit lamps etc)

The 32W lamps are not that far to cause any extra ill effect for the ballast. Rigorously you are right, but the ~2degC (= 2*3%*65degC) temperature difference really could not make anything. If the ballast have problems, it would have the same problems with the rated 36W lamp as well.

For the lamp it is a bit different story. The power dissipated in the tube of the same size is ~11% lower, so is the temperature rise from the ambient. On top of this the lamp is of completely different chemistry. So no wonder the operation conditions are restricted.
The temperature range was optimized for the range present in the majority of the present systems, not all systems. So in that majority it does save the ~3..3.5W of power input, what indeed may be quite a lot on the large installation with 100's of lamps.

I'm curious about the American F32T8 because they actually have a higher lamp voltage (~125v) than the 40w. The 34w are lower voltage (~75v) of course but the 32w would seem to work fine on switch start as long as the open voltage is sufficient to keep them lit. As a result of higher lamp voltage, the current is way less than 36 or 40w on reactive ballasts. Of course, they are designed for electronic ballasts with voltages of around 600v.

US F32T8 uses krypton-argon mix, bit different ratio then the European F36T8, so it have higher arc voltage. The real reason, why makers split the EU vs US specs for 4' lamp I dont know, my guess is, then the lower current/higher voltage lamp have lower electrode losses, but require higher OCV, what was not a problem in the US (as the 120V mains need the autotransformer anyway), but it would be a problem in the 230V EU (it would ask for way more losssy autotransformer vs simple, more efficient and alredy widely used series choke in the preheat circuit).

And to make a bit of mess on the globe, makers came with this "energy saver" so, it have the same designations, but totally different electrical spec as the already used US lamp...

But I would be curious, how (and if ever) would work the US F32T8 in the European F36T8 preheat circuit...