Or so TEPCO says. The company also says it was the worker's error.
From what I can figure from the TEPCO's handout for the press in English on June 23 (their English reads like Google translation...maybe it is), the problem was in the Skid No.4 of the cesium absorption. One valve was open which shouldn't have been, letting some of the water bypassing the cesium absorption vessels No. 2 and No.3.
So, will this account for 1/20 of desired performance by the Kurion system? TEPCO is running the test with this valve closed.