We report the serendipitous detection of a 0.2 L$^*$, Lyman-$alpha$ emitting galaxy at redshift 2.5 at an impact parameter of 50 kpc from a bright background QSO sightline. A high-resolution spectrum of the QSO reveals a partial Lyman-limit absorption system ($N_mathrm{HI}=10^{16.94pm0.10}$ cm$^{-2}$) with many associated metal absorption lines at the same redshift as the foreground galaxy. Using photoionization models that carefully treat measurement errors and marginalise over uncertainties in the shape and normalisation of the ionizing radiation spectrum, we derive the total hydrogen column density $N_mathrm{H}=10^{19.4pm0.3}$ cm$^{-2}$, and show that all the absorbing clouds are metal enriched, with $Z=0.1$-$0.6 Z_odot$. These metallicities and the systems large velocity width ($436$ km$,$s$^{-1}$) suggest the gas is produced by an outflowing wind. Using an expanding shell model we estimate a mass outflow rate of $sim5 M_odot,$yr$^{-1}$. Our photoionization model yields extremely small sizes ($<$100-500 pc) for the absorbing clouds, which we argue are typical of high column density absorbers in the circumgalactic medium (CGM). Given these small sizes and extreme kinematics, it is unclear how the clumps survive in the CGM without being destroyed by hydrodynamic instabilities. The small cloud sizes imply that even state-of-the-art cosmological simulations require more than a $1000$-fold improvement in mass resolution to resolve the hydrodynamics relevant for cool gas in the CGM.
Download