The LEP ring was closed on March 16^th. Security tests and equipment checks were made in the following days, and first beam was injected on March 29^th , some 5 days ahead of schedule. First collisions, at 45.6GeV per beam, were achieved on April 3^rd . The following week was devoted to Z⁰ production for experimental calibration, interleaved with commissioning of high energy operation. The requisite amount of calibration data had been delivered by April 11^th , following which attention turned to running at high energy.

The sole aim for the year was to deliver significant luminosity at the highest possible energies. The maximum operational energy depends on a number of different parameters:

Available accelerating RF voltage. Its evolution over the years is shown in Figure 1. It was increased firstly by installing additional RF cavities, and secondly by raising the accelerating gradient of the super-conducting RF cavities from 6 MV/m (design) to 7.4 MV/m.

Figure 1: Evolution of beam energy, nominal RF voltage (design gradient) and available RF voltage.

For a given RF voltage, the beam energy that can be achieved is determined, and the likelihood of keeping the beams is then at the mercy of the stability of the RF system. With such a huge system, running well above design, trips of one or more units are inevitable. With a total current over 4mA in the machine, trips were in fact frequent, with a mean time of around 20minutes. Therefore in order to produce significant amounts of luminosity, it was necessary to run with some safety margin. Furthermore, the total current in the machine was limited to around 5mA in the interests of RF stability.

The strategy adopted for most of the year was to go into physics with a safety margin of 2 klystron units. After about an hour the beam energy was increased, in physics, to an energy corresponding to a safety margin of 1 unit. After a further hour the energy was increased again, until no safety margin was left. Beams were then lost at the first trip.

Maximum horizontal beam size. The horizontal beam size s_x is proportional to beam energy E, the rms horizontal dispersion D_x^rms, the betatron function b_x and the horizontal damping partition number J_x:

The increase of horizontal beam size with energy results in lower luminosity and larger background in the experiments. This is counteracted with a high Q_x optics and an operational increase of J_x through an increase of the RF frequency. However, the increased J_x reduces both beam energy (longer orbit) and RF voltage overhead (larger energy spread). For maximum beam energy it is desirable to run with the largest s_x (lowest J_x) possible.

Average bending radius. The energy loss per turn is a function of beam energy E and average bending radius r. The average bending radius was changed operationally by using additional bending contributions from horizontal dipole correctors.

The LEP energy has been maximised in 2000 by optimising all of the above contributions. This resulted in many different physics energies throughout the year. The energy gain from 1999 (101 GeV) to 2000 (104.5 GeV) is analysed in Table 1.

Table 1: Contributions to the energy increase in 2000.

Operating the machine under these conditions implied shorter and hence more fills. While some 2600 physics fills had been made in the previous 11 years, 1400 more were made in 2000 alone. With refilling so frequent, further efforts to reduce the turnaround time contributed significantly to the integrated luminosity. The average time between physics fills, which was always well over 2 hours until 1997, was for the first time under 1 hour in 2000. The machine performance at different energies throughout the year is summarised in Table 2 and Figure 2.

Table 2: Delivered luminosities at high energy in 2000. The data is averaged over the four experiments.

Figure 2: Integrated luminosity versus the number of scheduled days of LEP operation for the different energy ranges. In total a luminosity of 233 pb^-1 was delivered to the four experiments.

From Figure 2 one sees that for the first half of the year the majority of delivered luminosity was at an energy between 102.5 and 103 GeV (2 klystron margin) and between 103 and 103.5 GeV (1 klystron margin). For the second half of the year, with more stability of the RF system, much more running was done with 1 klystron margin, resulting in nearly all the luminosity being delivered at an energy above 103 GeV in order to maximise the search for the Higgs boson. Also visible in Figure 2 is a short period in late July where LEP was operated with no safety margin at all, to deliver luminosity at the highest possible energy (above 104 GeV) for a chargino search.

The 5107 hours scheduled for physics were spent as follows;

• Experimental access 5.6%
• Downtime 7.5%
• Filling without coast 15.9%
• Filling with coast 25.9%
• Physics 41.6%
• Energy calibration 3.6%

As in 1999, flexible scheduling of machine studies and energy calibration in periods when high energy running was not possible helped to minimise the impact of various problems on the overall physics production. Never-the-less, the machine downtime of 7.5%, one of the lowest ever for LEP, was a remarkable achievement considering the mode of operation.