(UPDATE) According to @jaikoman who follows and tweets on every single TEPCO and NRA press conferences, TEPCO poured ice and dry ice, thinking ice would float, cooling the top layer of water, and dry ice would sink, cooling the bottom layer of water.
Well they need a Plan C. Dry ice pieces they pour were apparently too small, and they all floated. Water remain unfrozen, and TEPCO says they will know by mid August whether the operation will work.
TEPCO says by dumping ice and dry ice they can lower the temperature of the contaminated water in the trench to about 5 degrees Celsius, then they will be able to form a continuous ice plug.
Here I thought that water freezes at zero degree Celsius. As the whole world is seemingly going crazy afresh this July, maybe TEPCO is correct that water does freeze at 5 degrees Celsius.
From TEPCO's photos and videos library, 7/24/2014:
On July 24, 2014, TEPCO started the experiment of dumping ice into the Reactor 2 turbine building trench, trying to freeze highly contaminated water which has refused to freeze despite 3 months of freezing effort. Workers dumped only 2 tonnes of ice, or 4 bags with 500 kilograms of ice each.
Workers seem to be wearing vests, probably to shield ambient radiation. The location is the oceanside (east side) of the turbine building, where, according to the latest survey map by TEPCO as of July 8, 2014 (which I had a very hard time locating in TEPCO's updated site) the radiation level looks to be about 0.20 millisieverts (or 200 microsieverts) per hour. According to TEPCO, workers spent two and a half hours dumping 2 tonnes of ice using shovels.
Locations of the trenches filled with highly contaminated water (most likely from April/May 2011), and the locations in blue squares TEPCO wants to create ice plugs so that no water from the turbine buildings enters the trenches, from TEPCO's presentation to Nuclear Regulation Authority on 7/7/2014, when TEPCO disclosed that after three months of attempt, the water was still not frozen (English labels are by me):
Part of TEPCO's survey map (7/8/2014) showing ambient radiation levels, with "Shaft A" marked (by me) in red square:
So why isn't the water freezing? According to TEPCO's convoluted explanation to NRA on 7/7/2014, it is because of the fluctuation of water levels in the turbine building which creates water flow through the gaps created by the pipes that go through the turbine building walls. The flow was strong enough to disturb the freezing process, which TEPCO hadn't anticipated from the mock-up.
I do remember from January, I believe, a meeting at Nuclear Regulation Authority in which TEPCO and NRA commissioners discussed these ice plugs. Commissioner Fuketa openly questioned the efficacy of the scheme, asking TEPCO why they were planning to create a plug right outside the turbine building where lots of pipes are going through in a narrow space, as you can see even in TEPCO's simplified presentation to NRA on 7/7/2014 (English labels are by me) below.
The red rectangle right outside the turbine building is the ice plug to be created. The purple pipe in the diagram going down to the red rectangle is where workers were dumping ice.
Commissioner Fuketa also expressed doubt that it would ever freeze. I think he even asked what TEPCO's "Plan B" was, in case it would not freeze. TEPCO's answer was that it would freeze. (Watching this futile exchange live, I kept thinking, "Why can't they just pour concrete?")
Well the water didn't freeze. Nowhere close. TEPCO's measurement shows the temperature of part of the water which should have frozen is as high as 15 degrees Celsius, after 3 months of freezing.
So dumping ice and dry ice, then, is TEPCO's "Plan B". And ask the god of physics to look the other way and make water freeze at 5 degrees Celsius at sea level.
Ahhh good (bad) old days are back... when TEPCO used diaper polymers, saw dusts, shredded newspaper to try to stop the same highly contaminated water in the same set of trenches from pouring into the plant harbor.
It feels it was only yesterday.