I'm no expert, but

The COSCO Guangzhou is a really big cargo ship, relatively recently commissioned.  It is powered by a ~53,000HP two-stroke diesel engine. 

Oceangoing commercial vessels are subject to few, if any, environmental regulations, largely regulated by economic interest - nobody wants a leaky fuel tank cutting into their margins.  But could a public/private partnership provide an economically viable solution to pollution?

There is precedent for nuclear vessels, many American ships are powered by small reactors.  Four commercial ships have existed, of which one currently operates - the icebreaker Sevmorput clears paths for trans-arctic shipping.  The three others were either politically or economically nonviable research test-beds with a secondary commercial purpose.  Their failure was one of scale and focus, as Sevmorput shows.  The Japanese Mutsu had a leaky reactor which was caught in design but ignored in typical Japanese fashion (not typical, you say?  Take a look at the other nuclear incident in Japan, Fukushima, and the incompetent, self-serving coverup).  The NS Savannah, the first nuclear cargo ship, was just too small, and was designed with that in mind.  The Otto Hahn was a research ship, an inherently less-than-reliably-profitable enterprise.  But shipping and icebreaking are both reliable and profitable. 

Using the 53,000HP estimate, let's round the resulting KW from 39,500-something to 40,000KW.  This gives us a little wiggle room for electric motor inefficiencies.  So the total power draw is approximately 40MW. 
But we don't want to use the reactor at 100%, all the time.  OECD-NEA.org claims in a load balancing study (https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=13&ved=2ahUKEwibuYitg_viAhVvU98KHS_1A9sQFjAMegQICRAC&url=https%3A%2F%2Fwww.oecd-nea.org%2Fndd%2Freports%2F2011%2Fload-following-npp.pdf&usg=AOvVaw1yJ_QDWYRMY0eqJkJuz6ng) that the typical light water reactor can be easily ramped up or down based on demand.  Oceangoing military vessels use extremely high enrichment, producing so much power they have to be "poisoned" so their power generation stays constant over time instead of falling off a cliff.

My own research revealed a very useful study that I am totally unqualified to critique but which seems to be pretty good.  Thomas Dominic Ippolito Jr.'s EFFECTS OF VARIATION OF URANIUM ENRICHMENT
ON NUCLEAR SUBMARINE REACTOR DESIGN (1990) describes the space requirements of a seaborne, submersible nuclear reactor at low and high enrichment values, as low as 7% enrichment.  Unfortunately, the maximum civil enrichment is 5%, but there is good news: rather than the 40MW requirement we assume, his minimum is 50MW, a full 25% higher.  Can this be compensated for?  I bet it can.  After all, the 1963 NS Savannah produced 74MW despite its small size, all at less than 5% enrichment.

So why not use a nuclear reactor to power the propellers?  Likely politics.  Insurance is a nightmare that required a completely new law in the US.  But COSCO is a state-owned company, and the world's second-largest economy, already a nuclear power, could afford to take on such a political and economic risk.  It would certainly put other nations' shipping industries on notice about their own emissions and reduce port idle fuel consumption.  It might reduce turnaround.  It would increase reliability - reactors are unexpectedly simple in principle, and brushless or stepper electric motors don't deteriorate. 

For now, Russia leads the way.  I'm an unabashed nationalist, so that's a bit disappointing to me, but it's certainly a good motivator.  Good job, Russia.