The North Korean attack plan

Modern War Institute:

...
So what would the initial North Korean attack look like? Unlike in 1950, when Soviet, Chinese and North Korean planners had a realistic prospect of conquering South Korea in a war, an attack today has only a small chance of victory within a narrow time window. In 1950, DPRK forces enjoyed large initial advantages in numbers, troop quality, and all classes of equipment including tanks and aircraft. Today all of those initial advantages except artillery and numbers are on our side, and even the numbers are more balanced. DPRK planners recognize their inferiority in technology and, after ROK mobilization, even in numbers. They understand that ROK/US forces will have air superiority initially, and (unless China intervenes) air supremacy within days. They therefore plan to win by striking quickly, by surprise, while ROK forces are still mobilizing, US reinforcements are not yet in theatre, and while our airpower is largely committed to overcoming the DPRK integrated air defense system and targeting WMD storage sites, launchers, and command, control, and communications (C3) networks.

Recognizing that ROK forces will be on some degree of heightened readiness during a crisis, the regime will use its formidable intelligence and special operations capability to obscure preparations for an attack and slow ROK responses. Its own past history of symbolic attacks, placing its forces on alert, and angry promises to destroy its enemies will actually work in its favor in this case: ROK/US intelligence agencies will expect some kind of posturing from the North and may therefore misidentify attack preparations as lesser actions. DPRK agents will also count on the psychological reluctance of the South Korean population and government to believe that war is imminent. They will actively seek to influence the ROK democratic decision-making process to get inside our decision cycle. In particular, ROK mobilization will require a political decision and every hour of delay imposed through threats, deception, information and cyber attacks, or direct action will have consequences. In the end, even if ROK/US commanders do recognize the signs of an attack before it begins, it will still take time to react. In that time, DPRK commanders hope to win.

There will be no need for detailed orders. Just as ROK forces know and rehearse their war plan, DPRK forces are largely in place, in numbers sufficient to achieve some local breakthroughs on the major routes towards Seoul—their first operational objective. North Korea will hope to begin mobilization before South Korea does, and thereby turn their currently modest advantage in numbers into a temporarily significant one. DPRK forces will rely, Soviet-style, on the use of overwhelming artillery and rocket fires to break through ROK prepared positions along the DMZ, while using deep fires to attack C3 nodes, routes forward, and mobilization centers. Strikes against targets in Seoul and the surrounding urban areas will have the additional useful effect of causing fear and choking routes with a panicked populace.

On the subject of routes it is worth considering the limited space for mechanized maneuver in central Korea: The eastern half of the peninsula is largely mountainous with roads running along valley floors. The grain of the country will tend to push DPRK forces southwest (towards Seoul). The western half of the peninsula around Seoul and the Han River system is slightly flatter, but at least south of the DMZ the land is now so built up that once major routes come under fire it will be slow going for both sides. It’s not good country for heavy forces, and until recently both sides planned to use mostly lighter infantry to fight on the line. Recent announced changes to ROK force structure see a much greater emphasis on heavy forces—perhaps to get more combat power out of a smaller overall force—but the terrain suggests that such forces will likely be difficult to maneuver. Furthermore, DPRK tactics emphasize the use of infiltration to achieve local penetrations and attack deeper, tactical targets. Their line formations include elite sub-elements specially trained for these tasks, and the terrain—whether urban or forested mountain—is ideal for it. Road-bound heavy forces will be especially susceptible to such tactics.

The final element in the DPRK plan is an extensive deep battle across the entire South Korean depth using some one hundred thousand special operations forces (SOF). An interesting feature of this war is that since both sides look and speak more or less alike, covert insertion and operation is easier for each side—but especially so for North Korean agents who may move freely within South Korea’s open society.
...

There is much more.

Clearly, both sides have war-gamed the ground force maneuvers.  Friction will become a factor for both sides as the mechanics of such a large operation get bogged down by events large and small before the sides actually clash.  I suspect the US will have immediate air superiority and will move quickly to destroy the artillery positions that threaten Seoul.  Air superiority and satellite coverage will make it difficult for North Korea to mass its forces for the kind of attacks it launched in 1950.

That is why the Norks are expected to rely on special ops raids deep into South Korea.  A preemptive strike by the US and its coalition partners could target these units in their bases in the North.