Why is an electrostatic accelerator so inefficient at converting wall power to beam power? I was reading up on electrostatic particle accelerators when I read a statistic stating the efficiency of converting wall electrical power (the electrical power from the outlet) into beam power in an accelerator. I tried finding out why this was the case and doing some first cut calculations. The most I could come up with was a lot of the power gets wasted by particles not being efficiently collimated, I.E. most of the accelerated particles hit the wall. Now, I'm pretty sure that is not the only reason, I assume depending on the quality of the vacuum, scattering by the leftover gas molecules would also contribute to the lost beam power, as well as losses via emitted radiation from focusing the charged particles with magnetic fields. 
Am I getting this right? Or am I missing some huge power draw that sucks most of the power away from the beam? I know the efficiency depends somewhat on the beam maximum energy (since higher energy beams have higher brehmstrahlung and cyclotron radiation losses), I'm looking at the power efficiency of a 10 MeV electrostatic electron beam accelerator. 
Why are 10 MeV electron electrostatic particle accelerators so inefficient?  
 A: As one data point for linac efficiency, I found a nice presentation from a Jefferson Lab meeting (where CEBAF is - Continuous Electron Beam Accelerator Facility). This is from Googling 'energy efficient linac'. In it, they quote that the CEBAF klystrons are about 25-28% efficient, while their new solid state amplifier proposal (an SBIR, can see at Far-Tech, I have no affiliation other than Googling them) is 55% efficient.  This is at 1497 MHz, and 6.5kW linear mode amplification. 
Now, that ignores efficiencies in the electron source, focusing elements, steering plates, and all the vacuum components (a 300 liter/sec turbo plus backing runs about 700W). It also ignores production of cooling water and so forth, that will contribute to the overall (in)efficiency (and you will need lots of cooling of whatever target you are trying to get gammas out of).  We won't even go in to the electron-to-gamma conversion efficiency, much less how you are realistically going to extract energy from the gamma beam on the other end.
A: The 'electrostatic' accelerators don't recycle the current, I suspect.   This
would be a pelletron or Van de Graaf design, and current going through the free electron laser
magnet structure (undulator) then hits a target and dissipates all the energy
it didn't lose in radiation.   Hitting the target dumps all the kinetic energy.
A synchrotron, on the other  hand, takes bunches of relativistic
charges and moves them through that same undulator once per cycle, circulating
at near lightspeed, so a tiny amount of the beam kinetic energy can be lost
due to radiation reaction, but built back up by the accelerator sections 
elsewhere in the apparatus.   Those sections use timed RF in the right
phase to re-accelerate the bunches back to full speed, and until the
charges hit gas molecules and scatter into a beamtube wall, that 
beam can remain circulating for hours between fills.
A static field just makes a current from HV to ground.
Circulating is more efficient use of fast-moving particles than dumping
them into a grounded target; the synchrotron electron gun can be turned off after startup,
but the Van de Graaff has to put new electrons out continuously.
