Metropolitan Transport Information System
by Digital Geographic Research Corporation
25+ years innovating

Turn Time — Meltdown or Summer Heat Wave?

2014 11 12 | Val Noronha, PhD

Revision history:
2014 11 13 | Conclusions clarified with r-squared statistic
2014 11 18 | Epilogue: October volumes

Truck turn time in Los Angeles-Long Beach hit another record in October, with 36% of all trips over 2 hours, and a new low of 31 on the 100-point METRIS Performance Index. In 2010 the time a trucker needed to budget for a trip to meet a deadline was 2-3 hours. This past July it was 3-4 hours. In October it was 4-4½ hours.

If there's a surprise, it's that the numbers weren't worse. Alarms are going off practically every day about this and that statistic in the red — like the paragraph above — and the term “meltdown” has been floated.

The Journal of Commerce (JOC) lists 12 reasons for current congestion. They're all well reasoned. That doesn't mean they're all having a measurable impact, or that they all need to be addressed today. Which of the 12 are most responsible for what's going on now? How are the solutions to be prioritized? As is often suggested, the answers lie in good data and analysis.

The data indicate that what's happened in the past few months is that the ports have reached their saturation capacity. A greatly diminished capacity compared with even the recent past. But there have been no new significant generators of delay for 6 months, and current performance issues are attributable almost entirely to container volume. Bear in mind that September trade was the 4th highest on record.

This is good news. Because performance degradation under a known stress is predictable and manageable. When that stress is volume, degradation is also reasonable.

What Does It Mean to Have Reached Saturation Capacity?

Think about a city freeway at 3 a.m., with 2 vehicles per minute in each lane. We're flying along at 65 mph. Double the traffic to 4 per minute, 8/min, even 16/min and there's no drop in speed. There's capacity to spare. When lane density reaches 30/min, things get touchy. Add more vehicles and speed slows; take them away and it recovers. So at this point there's a causative relationship between vehicle volume and traffic speed. The freeway is exhibiting saturation capacity behavior. It can absorb more traffic, but with a performance penalty.

Now if it were to snow, cars would have to maintain a greater headway distance, and saturation capacity behavior would begin at 20/min rather than 30/min.

If speed and headway were to decline in the absence of additional traffic, we'd suspect a complication: an obstruction, perhaps an accident. At 2 vehicles/min the obstruction would not have caused the same degree of slowing. So there's insidious value in operating near saturation capacity.

Saturation Capacity in Los Angeles-Long Beach

The graph in Figure 1 shows the METRIS Turn Time Performance Index (blue) and container volume (gray) in San Pedro over the past five years. There's no apparent relationship between the two. Logically, turn time should worsen (i.e. Performance Index should decrease) when traffic increases. Why does that not happen? Because times have been good — like the freeway at 3 a.m. When volume increases, terminals hire more labor (Figure 2). There are reasons why truck turn time performance has been in a steady decline for years, but volume isn't one of them.

Figure 1. The mirror image that develops between the two curves starting in spring 2014 suggests that the ports are currently operating with a diminished capacity of around 1.2 m TEUs/mo, beyond which volume is directly reflected in poorer truck turn time performance. Prior to 2014 that was not the case because there was copious surplus capacity. Mouseover/click the figure to see the drop in capacity.

Figure 2. Labor hours per TEU have been steady for years, indicating that terminals hire more labor during periods of high demand. Labor inputs have not been wanting during the current performance dip.

Until 2013, swells and dips in monthly volume were routinely absorbed by the system. While the Performance Index generally declined, it remained in the 45-60 region. Even the 3-year monthly high of 1.34 million TEUs in August 2013 was a non-event in turn time impact. There was still capacity to spare.

The latest data (September container volume, October turn time) have put enough points on the graph to create a pretty convincing case that the port is in saturation capacity mode. It's been in this mode since the spring. Volume didn't increase in spring; capacity had plummeted over a period of time, to 1.2 m TEUs/mo, due to the convergence of chassis issues, mega-vessels … those 12 factors referenced by JOC. Starting in the spring, any volume over 1.2 got perfectly reflected in turn time. Unsurprisingly, the 1.4 m TEUs in September, the 4th highest month ever at the ports, caused a precipitous drop in the performance index.

A second observation from the graph is that capacity may have bottomed out at about 1.2 m TEUs/mo — the perfect reflection suggests that it has not fallen further in the past several months.

Conclusions and Caveats

METRIS analysis and consulting help MTOs, LMCs, BCOs and ports achieve consensus.

Current problems are severe, and the loss of capacity is intolerable. The opening paragraph above is unambiguous, and the conclusions below take nothing away from that.

But looking at the causes of increased congestion, the constant alarms, and the distraction they constitute for management, the finding that arises from the analysis is that nothing significant has just happened, that hadn't happened 6 months ago. The recommendation is to maintain a steady course: crank up capacity by fixing the strategic issues that were strategic issues 6 months ago (e.g. chassis), but there's no reason to second-guess directions every week just because of new reports of falling performance. This is largely about volume. For the record, in statistical terms, volume accounts for 86% of the April-September variance in turn time (though with just 6 points, that calls for a pinch of salt). When/if volume falls back to 1.2 m TEUs/mo at the end of this peak season, turn time and everything else will stabilize. For a month. There is no hidden agenda in this recommendation; it is based entirely on the data, and it should be read in the context of the limitations of the analysis.

This is one way of interpreting the data. There are others; the capacity curve could be drawn in different ways, and more sophisticated analyses are possible, e.g. the impact of dual roadability inspections can be established quite precisely. Saturation capacity behavior is recent — six months does not present enough of a history to form firm conclusions. Nevertheless, the graphic evidence (sudden onset of mirror-effects in the curves) coupled with basic transportation science, is remarkable and persuasive. This analysis is based on observations on truck turn time only, not container dwell time (ship-to-exit-gate) which has also fallen out of control, but is probably a domino effect.

An unmistakable takeaway is that there are data that can illuminate what's going on, and analysis is key to understanding and remedying the congestion problem.


Epilogue — Something Grim Happened in October

2014 11 18

The article above was posted on November 12, based on container volumes to September, and turn time to October. On November 17, the ports released October traffic statistics.

The figures were low. The reasons are fairly obvious: the season is tapering down; vessels are anchored offshore instead of discharging cargo; and with San Pedro choking, other ports are beckoning the traffic.

But that’s not all of the bad news.

In Figure 1 above, the pattern of reflection was perfect, assuring. Although October turn time performance was at a record low, it was a continuation of a 6-month pattern, reasonably suggesting a strain against volume. The figures below, which incorporate October volumes, tell a troubling story.

Figure 3. The same graphs as in Figures 1 and 2, with October volumes included. Left: Despite relief on container traffic, performance sank further. Right: Symmetry emerged in labor hours vs turn time. For the past two years, the pain of MTOs and LMCs has been in a near-perfect pas de deux.

The bad news in Figure 3 is that the October numbers blew the symmetry we saw in Figure 1. We can not take comfort in the notion that the performance dip in October was all about traffic, as it was in April-September. Performance plunging to a new low, despite a volume decrease, is a sure sign that something else was going very wrong. The METRIS index was 10 points below what we'd expect for that volume. Ten points are 1½-2 times the gross impact of the 2012 strike (the strike lasted 8 calendar days; the effects here are over a 30-day period, so the impact is less than that of the strike on a per-day basis, but still substantial). The pain was intense for both terminals and truckers.

I humbly withdraw the contention that nothing new has happened in 6 months. That was true as far as September. Something new did break down in October, and it's important to get to the bottom of it, identify and measure its components, and address them. Identifying the components is not my specialty. Based on the data and reputable news sources, I'd guess at one or both of:

  1. An entirely new source of delay. Labor problems come to mind, particularly the reported secondary roadability inspections and equipment staffing. Fortunately it is possible to quantify some of this delay. It is also possible to address it relatively quickly, at least in theory, with the right will and human factors.
  2. An existing source of delay either echoing (e.g. backlog), or spiraling out of control due to domino effects. Examples: (a) As turn time worsens, containers don’t get picked up on time, they remain on the dock, increasing stack occupancy, which slows turn time further. (b) Containers arriving off-schedule at BCOs take longer to process, delaying backhaul of empties. (c) Deadhead trips (bobtails, bare chassis and empties) increase. (d) Chassis shortages lead to hoarding by BCOs and MTOs, compounding that problem. These are difficult to isolate and measure, and it will take more than will to fix them. They require an extraordinary application of corrective force, at some cost, to arrest the spiral.

At least three commendable fixes are under way. Terminals are flexing breaks, and the effects are noticeable particularly on 10 pm queues. A chassis agreement promises to ease the chassis problem after February 1. An off-dock empty storage dump will relieve terminal congestion.

A respite from record traffic is on the horizon, at least for a few weeks.

Will these suffice?


“U.S. Container Port Congestion & Related International Supply Chain Issues: Causes, Consequences & Challenges.” Federal Maritime Commission, Bureau of Trade Analysis, 2015 07 14
The statement on page 54, attributed to a METRIS report: “[evening queues are] not being addressed because the marine terminals that operate the PierPASS program do not consider the long lines outside the gates as their problem....” is not taken from DGRC's works.

“Near-record volume is main driver of LA-LB congestion, study finds.” Journal of Commerce, 2014 11 14

“LA-Long Beach congestion swells on longer container dwell times.” Journal of Commerce, 2014 11 24

“Increasing container dwell times are taking a toll.” CFR Rinkens

“LA-LB truck turn times worst ever, but long box dwell time worst of all.” Shanghai Shipping Exchange, 2014 11 28 Turkey Sea News, 2014 11 28

“Chicago drayage operators worried port congestion could hurt their business.” Livingston


The METRIS data archive, on which this article is based, was possible due to the cooperation and patronage of leading drayage carriers who have been our partners over the past 7 years: Ability/Tri-Modal, California Intermodal Associates, Dependable Highway Express, Fox Transportation, G&D Transportation, Golden State Express, Harbor Division Inc, LA Grain, Port Logistics Group, Price Transfer, Progressive Transportation Services, Southern Counties Express, TK Transport, TTSI and Yamko Truck Lines.


Add Your Comment


E-mail (will not be displayed)

Organization and Location

Comment (700 chars left)


© Digital Geographic Research Corporation