Hurricane Beryl reached Category 5 much earlier in the Atlantic hurricane season than any other system on record. On July 1, as a CAT4, it devastated Carriacou. Jamaica’s south shore received a harsh passing swipe. On July 5 Beryl made landfall at Tulum as a high-end CAT2, continued across the Yucatan, then entered the Gulf of Mexico as a tropical storm. It was a strengthening CAT1 hurricane as it came ashore at Matagorda, Texas on July 8 (more and more).
Before dawn, eastern time, on Saturday, July 6, I sent the following note to clients, colleagues, and friends:
This morning our cone of uncertainty extends from the mouth of the Rio Grande to Houston. Until we are more certain of how much force on which specific targets — we can ask ourselves questions, preparing our minds (and more) for a range of potential answers.
There will be lots of rain where there has recently already been lots of rain, so there will be flooding. Bad enough to bring down bridges and stop trucks? Bad enough to swamp water treatment facilities? Are major grocery distribution centers and such at risk? Here’s one place to watch and to find other places to watch.
There will be storm surge. Bad enough to disrupt LNG infrastructure? (Here and here and here) How about other energy nodes and channels (here and here)? Here are a couple of places to watch: Enbridge Corpus Christi and Aransas Pass.
Starting Sunday night there will be tropical force winds. We can be almost certain of disrupted local electrical distribution networks. Transmission networks? Generation facilities? Will the grid-free zone be sufficiently narrow to support continued inbound trucking? If the grid is gone for several days for tens-of-thousands, the ten day forecast looks dangerously hot with even nighttime temperatures at 80 or above.
Right now the forecast does not support my worst-case worries. Right now it looks like the storm will avoid the largest concentration of migrants without shelter. Right now grocery DCs are likely to survive, the roads serving most places will be fine, and refueling should be possible. But we also know that late stage intensification is a growing problem. So, I am mostly scanning for capacity concentrations while waiting for both force and targets to be clarified.
I was recently reminded that reconnaissance is meant to help us “recognize” — rethink — reality. Beryl may yet cause me to recalculate what I think I know: adapt, adapt, adapt…
One week later how does that rough two-days-ahead risk assessment (Threat x or / Vulnerability) hold up?
Late yesterday S&P Global reported, “Almost 1 million electric customers remained offline in East Texas July 12, five days after Hurricane Beryl barreled into the Gulf Coast, and about 500,000 customers may remain in the dark July 15 amid a continuing National Weather Service heat advisory.” In the three days after Beryl’s passing daytime temperatures in Houston were above 90 Fahrenheit with dewpoints over 70. All HEB stores and most other grocery retailers were open on Tuesday. Public water disruptions have been minimal. Retail fuel has continued to flow. The big Marathon and Valero refineries are operating. Fuel racks are dispensing product. Fuel tanker trucks are delivering. The port of Houston has reopened. While the electrical distribution system was left in tatters, the transmission network and generating capacity remains intact. In fact, ERCOT has had surplus capacity because demand was suppressed by lost connections.
On Tuesday there was the typical demand surge for gasoline that drained several retailers. Many stations without back-up power were closed. Much of Wednesday and Thursday was spent refueling sold-out retailers. There are still several gas stations without power to pump. But there are even more that are open for business (more and more) I have not (yet) found a major truck stop that is closed. Yesterday (Friday) the Freeport LNG terminal had not returned to full operations, but there were widespread expectations that it would restart late yesterday or this weekend. Early estimates of insured losses in Texas are running above $2.5 billion (more).
There have been many supply chain challenges. Anytime the grid is gone there will be serious challenges. When the grid is gone for an extended period is exactly when Supply Chain Resilience is tested. In my judgment, the metro-Houston supply chain has passed this (early-in-the-semester) test. This does not diminish continuing human (and economic) consequences of flooding, wind damage, storm surge, and grid failure. But in the aftermath of a hard hit, water, food, fuel, and other crucial freight has continued to flow at velocity and volumes sufficient to supply millions. The grid-loss has been very tough, but consider how bad it would be if amplified by the absence of public water and commercial grocery and fuel operations.
In retrospect there were two errors in judgment — implicit but influential — in my July 6 risk assessment. I failed to consider the increased grid vulnerability resulting from mid-May’s derecho in the Houston area (more and more). This almost certainly pre-set ground and tree conditions that explain some portion of Beryl’s impact. Tree roots that were weakened in May gave way. Repairs not quite completed since May were even more vulnerable to the hurricane. Recurrence is a key principle of risk assessments. I did not give this principle enough attention. My expectations for a slightly more westerly track by Beryl — not quite as Houston-specific — did not fully adapt to updated weather forecasts. As the map below suggests, by Sunday morning a bulls-eye on the urban matrix was reasonably certain. But my attention was distracted by my own expectations and lack of concentration (on Sunday I was engaged in many non-Beryl personal activities). Combined with neglecting derecho impacts, this distraction resulted in a level of grid loss that surprised me, despite plenty of evidence that could have generated more accurate anticipation. In this particular case, improved anticipation would not have resulted in different actions. But in other contexts, accurate anticipation can be the best action advantage available.
No new storm in the Atlantic this morning. Beryl was a helpful prompt to pay closer attention next time.
+++
July 16 Update: As of late Monday afternoon, roughly 200,000 Houston area electric customers remained off-the-grid. Inbound deliveries of feedgas to the Freeport LNG facility resumed over the weekend. But outbound train-flow is not yet moving. A Freeport LNG spokesman said they are repairing, “damage sustained to our fin fan air coolers in the hurricane and anticipate restarting the first train this week.” S&P Global explains, “Fin fan air coolers are used to dissipate heat at liquefaction plants. They need ample air supply to operate, so they are often more exposed to weather events than other pieces of equipment.” US natural gas prices have dipped a bit on the loss of Freeport’s demand. The Dutch TTF benchmark was lower on Monday and today, Tuesday, has opened trading in a narrow range. Europe’s natural gas storage domes are already just about full for winter (here). July 17 related piece by Bloomberg, “Gas futures for next-month delivery sank 7% to settle at $2.035 per million British thermal units on Wednesday. October futures slid 5.3%, while the January contract dropped 2.3%. The outage at the Freeport liquefied natural gas terminal has dented consumption and US data continues to show larger-than-expected weekly additions to “already bloated inventories…”
July 20 Update: According to the Houston Chronicle, “The Texas Windstorm Insurance Association (TWIA) said in a meeting this week that it anticipates draining half of its catastrophe reserve funds to cover windstorm damage payouts. The TWIA is an insurer of last resort for many Texans, giving policies only for wind and hail damage to customers who have been denied coverage in the public market. Chief Actuary Jim Murphy said that the agency expects to use about half of its over $450 million catastrophe fund to pay for the 16,000 claims it has already received. That figure could go up to 20,000 claims for an estimated more than $200 million in payouts. ” According to Reuters, Freeport LNG is receiving inbound feedgas, but as of Friday had not yet begun loading at least six LNG tankers anchored nearby.
July 26 Update: S&P Global reports on CenterPoint’s plans to do better next time. Freeport LNG outbound flows restarted on July 22. Estimates of US insured losses now seem to be settling into a range of $2.5 billion to $4.5 billion, with another $2 billion or so related to insured damages caused before the Texas landfall. Since shortly after Beryl’s transit near record Sahara dust storms have suppressed formation of new Atlantic hurricanes. This effect is not atypical for late July and usually dissipates by mid-August.