Pedestals limited by peeling instabilities have been reached in JET-ILW by operating at high q95, up to q95=8.5. The increase in q95 via the increase of the toroidal field has stabilized the ballooning modes and has allowed to reach high pedestal temperature (up to 1.5 keV for the electrons and up to 2.2 keV for the ions) and low pedestal density ( approximate to 1.8x1019m-3), with electron-electron pedestal collisionality approximately 0.15 and normalized ion Larmor radius 0.002 approaching ITER normalized pedestal parameters. The most unstable pedestal instabilities are peeling with toroidal mode numbers in the range n=1-5. A density scan in peeling limited pedestals shows that the increase of the pedestal density leads to an increase in the pedestal pressure. The modeling shows that this effect is due to the stabilizing effect of the density on the peeling modes. On the contrary, the increase of the separatrix density does not seem to affect the pedestal pressure in peeling limited plasmas. These behaviors are opposite to those observed in ballooning limited pedestals. An isotope mass scan from pure deuterium to tritium-rich plasmas has been performed with peeling limited pedestals. The increase of the isotope mass leads to an increase of the density at the pedestal top, via the increase of the density gradient. This behavior is similar to that observed in ballooning limited pedestals. The increase of the isotope mass also leads to the increase of the pressure at the pedestal top, via the increase in the pressure gradient. The temperature is not affected significantly. The increase in the pressure is not ascribed to a direct effect of the isotope mass on the pedestal stability, but to an indirect effect due to the increase of the pedestal density which, as shown in the deuterium density scan, stabilizes the peeling modes. The experimental results are used to validate the pedestal predictions using the Europed code. In all the scans performed, a good qualitative agreement is observed between the predictions and the experimental results. Quantitative disagreements can be in part ascribed to the fact that a consistent modeling should integrate the effect of core and scrape-off layer.