Bill Gretta, principal, SCR Solutions LLC, presented two case studies at a recent HRSG Forum virtual session, in which a unique field test method combined with sophisticated CFD analysis suggested modifications for improving distribution of ammonia through the SCR catalyst modules to improve NOx reduction and ammonia slip. Old units, and many new ones, are not equipped with a permanent NH3 sampling grid downstream of the SCR, and it’s costly to add, Gretta said.

He described a method that makes use of a flexible weighted probe with NOx and NH3 sensors which is lowered into the SCR inlet and exhaust gas flow fields from multiple ports on the roof (photo). Then EPA Test Method 320 is applied. Subsequent CFD analysis revealed the reasons for areas of high and low NH3 concentrations after the ammonia injection grid (AIG).

In the first case study, a 2 × 1 501F-powered combined cycle with close to 500-MW output, this approach resulted in removing and rebuilding the AIG, locating it three feet closer to the CO catalyst, and adding mixing baffles and plates to reduce the root mean square (RMS, an indication of the quality of distribution, the deviation from the average of many values) of ammonia-slip variance from 70% to 10%; additional tuning got it down to 6%. Buildup of ammonium bisulfate in zones of high ammonia slip decreased dramatically.

In the second case study, a 2 × 1 501D-powered combined cycle installed more than 25 years ago had to meet a lower emissions profile, so a dual-function catalyst was selected, but failed to meet the new standards. Analysis showed there was plenty of catalyst, so other system issues were at play.

Gretta and his team simulated 501D exhaust, sampled at 50 data points in a 5 × 10 array of SCR inlet and outlet locations with the weighted probe, and then did an inspection and CFD modeling when an RMS value of 19.3 indicated poor distribution. Causes of poor distribution and solutions were similar to those identified in the first case study.

Insights gleaned from the Q&A included the following:

• • In both case studies, AIG heavy support elements (which Gretta said probably would be found only in early SCRs) were getting in the way of flow; the replacement was designed to be self-supporting to eliminate the old support structures.
  • Rust and scale were blocking the AIG ports. The new AIG uses stainless steel instead of carbon-steel pipe and includes cleaning and vacuuming ports in each lance. Hole diameters also were increased and rearranged.

UPCOMING HRSG FORUM EVENTS

During the HRSG Forum’s second monthly meeting, June 2, 2021, hosted by Bob Anderson and Barry Dooley, close to 130 owner/operator representatives from 34 countries (out of 219 total attendees), were enlightened on two vexing issues with HRSGs: (1) steam-side oxide growth and exfoliation (OGE) from superheater (SH) and reheater (RH) tubes, and (2) the use of computational fluid dynamics (CFD) and field testing to improve selective catalytic reduction (SCR) unit performance.

Judging from the number and quality of the questions for both presenters, these attendees weren’t just staring at their screens. You don’t want to miss listening to recordings of the presentations all available at HRSGforum.com. They are rich in detail with a methodical sequence of illustrations for truly understanding the problems, impacts, and solutions.

Barry Dooley, a senior associate at Structural Integrity Associates Inc, whose experience dates back decades to some of the early work done at the UK’s CEGB, Ontario Hydro, and EPRI on OGE, explained how SH and RH ferritic steels like T11, T22, T5, T9, T23, and T91 are susceptible to oxide growth on inner surfaces containing greater amounts of hematite versus magnetite, which can lead to exfoliation of particles under the right thermal stresses (Fig 1). The progression of formation for different alloys, from laminations in the oxide layer to cracks to exfoliation, is well depicted in the slides.

The varying alloy compositions—chromium and molybdenum contents specifically—help determine how fast deposits grow, and the risk of exfoliation. The specific environmental factors are saturated or superheated steam, gas-turbine exhaust temperatures from 1100F to 1150F, use of duct burners, and tube temperatures ranging up to 1200F.

The deposits themselves can lead to tubes operating at higher temperatures, resulting in an ever-increasing oxide growth rate. The exfoliated material causes erosion, plugging, and sticking in valves; erosion of downstream HP and IP steam-turbine inlet-valve and steam-path components; or simply collects in a header (Fig 2). Impacts tend to show up after many thousands of operating hours but of course are aggravated by deep unit cycling and starts/stops, once the oxide reaches the critical thickness for exfoliation. Dooley shows one HRSG case in which material began to exfoliate after only 24,000 operating hours.

Unfortunately, OGE cannot be controlled through steam/water chemistry changes. It’s not dependent on O2 concentrations, but instead on O2 partial pressure. The influence of film-forming substances in the chemistry is uncertain. Shot-peened 304H and S304H SH tube alloys will exhibit a Cr-rich layer along the surface which slows the rate of exfoliation. “It’s rare for them to exfoliate,” Dooley said.

Among the insights that emerged from the Q&A session:

• • No relationship has been developed among operating parameters (for example., total operating hours, number of starts, etc) and OGE to predict its onset before impacts occur.
  • Cycle modifications which increase gas-turbine exhaust temperature raise the risk of oxide growth.
  • UT analysis can detect oxide-scale thickness but only lab metallographic analysis can reveal the characteristics of the oxide layer critical to OGE.
  • Early theorists suspected that steam/water O2 levels contributed to hematite formation, but deeper research has proved this false.
  • Small additions to the alloy, like vanadium and tungsten, will alter iron-ion migration patterns.

UPCOMING HRSG FORUM EVENTS

If the CCUG 2021 Day One roundtable on cold-weather preparation is any guide, winterization continues to vex plant personnel. Many of the issues can be traced to inadequate design bases and insufficient equipment. However, the root cause appears to be building “outdoor” facilities in locations which clearly require far better protection against protracted frigid conditions. Exhibit One: The tragic consequences in Ercot this past February.

For this virtual roundtable, convened in the early afternoon (Eastern time) of July 13, Steve Hilger, plant manager at Dogwood Energy (operated by NAES Corp) and a member of the Combined Cycle User Group’s steering committee, was joined by Mike Armstrong, engineering manager at Competitive Power Ventures’ Woodbridge Energy Center. Hilger’s plant is located south of Kansas City, Armstrong’s 2016-vintage combined cycle is in New Jersey about 20 miles south of New York City and near the shore. Both were designed as outdoor facilities.

Hilger’s first slide notes that the lowest ambient temperature experienced at Dogwood was -23F while the heat-trace design basis was +2F. Dogwood is two decades old and climate disruption is real, but surely this differential is better explained by designer negligence.

Woodbridge, says Armstrong, was designed to -8F but does not account for wind chill. “Most equipment, even our HRSG drums, lack enclosures, and are open to the wind.” Snow breaks and wedges were never installed on the roofs of outbuildings at Woodbridge. “Once we covered 175 valve handles to prepare for a cold weather event,” he lamented (Sidebar). Woodbridge has also added warming sheds on the top of the HRSG to keep personnel warm, and purchased several 120-Vac instrument space heaters wired to plug into outlets. These are used in transmitter boxes and ductwork with failed heaters.

Dig deeper: Heat tracing demands constant attention

Some of these temporary measures have their risks. Portable gas or propane heaters, for example, used in enclosed space elevate CO exposure to workers and present fire hazards.

The impacts aren’t solely on the equipment either. “Operators find reasons to be absent when it is cold outside,” Armstrong said. Woodbridge purchased steel-toed buck boots for winter work and makes space available in nearby motels to keep employees off the highways. We also remind them to look up for icicles in areas which may have leaks, he said, but also review work orders to identify equipment which may be leaking.

Equipment of special interest are drip-pot drains, attemperators, air filters (which can become plugged with ice), and outside air compressors, the last “a problem” at Woodbridge. “We had our heat-tracing contractor do an audit to compare design conditions to actual,” Armstrong said.

An audience member suggested that plants check for clogged drain lines using NDE. He said it takes his plant about three hours to check all HRSG drains in this way. And actions taken in other seasons, like a contractor removing heat tracing for a valve repair in summer, need to be checked.

Users struggling with winterization might spend quality time with the slides, which could be turned into a poster titled “Winter is coming” and hung in the plant’s lunch or common area. Here are a few bullet points:

• Review the 32F action plan (or prepare one if your plant doesn’t have).
• Review alarm points and operational permissives which may be impacted by cold-weather operation.
• Order bulk chemicals and review chemical properties to determine freeze points.
• Stage electric and propane heaters in problem areas.
• Develop HRSG drain procedures with valve identification in case there is only enough fuel to operate one unit.
• Perform a heat-trace audit in August/September and evaluate heat-trace insulation for deficiencies, keeping in mind that heat tracing cannot protect areas with water-soaked insulation.
• Verify calibration of all transmitters suspected of freezing or of over-heating.

With winter just over the horizon, the roundtable on winterization at the 2021 conference of the Combined Cycle Users Group was perfectly timed. There are still a few weeks, depending on your plant’s location, to implement best practices shared by colleagues from Woodbridge Energy Center, Dogwood Energy, and others.

By virtue of its location and importance to the grid, Woodbridge, a 2 × 1 7FA.05-powered combined cycle located outdoors in the Northeast, has heat-trace experience beyond that of many others in the industry. Engineering Manager Mike Armstrong represented his plant in the roundtable.

What follows are details on Woodbridge’s heat-trace initiatives, some of which were not discussed during the roundtable because of time constraints.

Get off on the right foot. Plant personnel learned during commissioning, and afterwards, that poor installation practices coupled with the lack of documentation made it difficult to troubleshoot the heat-trace system. This required staff to spend roughly 60 man-hours per week identifying and fixing issues with heat-trace circuits not functioning as designed. The poor performance of the heat-trace system jeopardized reliability and operability by allowing critical instruments and equipment to freeze-up.

Woodbridge was constructed by a single EPC contractor with multiple equipment suppliers. Design of the heat-trace system was subcontracted to a reputable supplier while installation was handled by the EPC contractor’s craft electricians, who had little or no experience with heat-trace equipment.

The various scope-of-supply boundaries and types of heat tracing proved problematic. Many field changes were required to complete the installation—changes performed without the knowledge of the designer and poorly documented.

Heat tracing was designed to maintain an equipment temperature of 40F at an ambient of -8F. The heat-trace supplier implemented the use of microprocessor-based temperature control and monitoring panels which required other new equipment—including various temperature sensors, new alarm capability, DCS integration, self-testing circuit cards, and programmable RTD outputs.

The lack of qualified oversite from the heat-trace designer during equipment installation and in preparing documentation of as-built conditions proved challenging for the plant operator once it took possession of the facility.

Dig deeper: Grid meltdown forces cold-weather prep to front burner

The first step in fixing the problem was to bring back the original heat-trace designer to audit the entire system and identify and correct any deficiencies. This required all 612 individual circuits to be reviewed to ensure the correct materials were used along with the correct installation practices. Next, all the documentation was updated to reflect as-built conditions. This information and a thorough review ensured the system was designed and installed as originally intended.

Why would NAES Corp, the largest third-party O&M services provider to combined-cycle facilities, run a three-day conference on NERC-related issues and compliance for its people and select clients and industry stakeholders? For one thing, NAES has 130 plants under its purview and supports over 100 plants outside its portfolio.

Then you might ask: Why does NAES have over 20 subject-matter experts (SMEs) in its NERC Compliance Services Group, which also includes a Compliance Testing arm? The answer: Managing NERC compliance risks for clients is a business in itself, and a key area of growth for the company. It is also considered a competitive distinction by NAES executives.

As one presenter, from a major ISO, put it, “the operations staff does not have the desire, or time, to review or interpret regulatory requirements in real time.” Put another way by another presenter, CIP (critical infrastructure protection) standards are now too complicated for the typical plant employee; many asset managers simply don’t have the experience or the background to manage NERC compliance risks.

A NAES NERC services group leader added, “NERC now has to be taken into account for any project, big or small, that has potential to impact facility output.” Scenarios he gave as examples include engine (GT) tuning, for which he listed six impacts of NERC standards; relay upgrades, which touch at least four NERC standards; and excitation control upgrades, which touch at least seven standards.

While the depth and breadth of the material presented was, well, almost overwhelming, much of it of most interest to combined-cycle owner/operators centered on EOP-011-2, recently adopted by NERC (but still in draft) addressing emergency operations and preparedness generally, and cold-weather readiness in particular.

One presenter suggested that valid bids into ISOs will be required to submit new weather-related data, such as minimum design temperatures, historical operating temperatures, and current cold-weather performance temperatures as determined by engineering analysis. EOP-011-2 includes deadlines and procedures for cold-weather preparation plans, identifying and tasking a responsible coordinator, definition of the “cold-weather period,” assessment of winter readiness, and new monitoring requirements for detecting failures, especially in heat-trace systems.

Update: There is now a Standard Authorization Request (SAR), entitled “Extreme Cold Weather Grid Operations, Preparedness, and Coordination” (Oct 6, 2021), that would add some additional requirements to EOP-011. These would include requiring annual training for generator owners and operators, detailed root-cause analyses for any freeze-related failures, and retrofit of existing generating units to operate in “extreme” weather conditions for their locations.

As one concrete impact, additional operator rounds will be required when ambient temperatures fall below certain limits, such as rounds once per shift when temperature falls below 32F, two rounds per shift when below 20F, and so on. Anything that could initiate an automatic trip should be included on these rounds. Training in cold-weather operations will have to be provided to site staff.

Reflecting, one highly experienced industry engineer advised that temperature is only part of the equation. Wind direction and velocity also are meaningful factors, he said, as are duration and current plant state (running, offline, etc). He recalled working at a facility that could run/start/operate at -20F when there was no wind. But when there was a 20-mph wind from the West it struggled at +20F. After five years of operation, staff was still ill at ease when confronted with uncommon temperature and wind combinations.

One of the presentations provided some instructive real-world cold-weather horror stories, including these:

• A 2 × 1 combined cycle in Nevada, designed with a 15F heat-trace system, experienced 7F ambient. The GT inlet bleed valves started behaving erratically, and operators were cycling through fans to keep the air-cooled condenser operating. When the cold-reheat pressure transmitter froze, it tripped the plant. One week later, the plant was able to restart.
• A 2 × 1 plant in Washington state suffered a cold snap and heavy snow. Natural-gas demand in the area overwhelmed supply lines. Ice and icicles formed. The fuel-gas system alarmed on low pressure. When the third fuel regulation station failed to operate, the unit tripped. Moisture trapped in piping froze.
• At a third facility, $100,000 worth of transmitters had to be replaced after a cold-weather event; water buildup in low points of oil feed pipes froze, and ice built up on the cooling towers because they lacked drift eliminators.

The presenter concluded with the mention of a plant which had to spend $4.5-million to be capable of running during extreme cold weather.

Among the other jewels of knowledge shared at the conference:

• Lessons at the industry level tend to get lost rather than learned. For example, Ercot experienced catastrophic cold-weather events in 1988 and 2011, and then again this past winter.
• Texas has 190 registered “generation owners”; 100 of them are new in the last five years, which means they may not be battle-tested for extreme weather events.
• The 70-member North American Generator Forum (NAGF) focuses on the NERC activity which directly affects the generator segment, and has six working groups, one of which focuses on the emerging cold-weather standard.
• NERC may create a “resilience” standard (in contrast to reliability). NAGF is busy determining the “cost of resilience,” a value to base investment on, since, according to the presenter from the group, organized markets do not support the value of resilience and the costs are born unequally among stakeholders.

You can spend a lot of quality time pointing fingers towards whom or what was to blame for the cold-weather issues experienced in several electricity markets—most notably Ercot—during February 2021. But no question, many gas-fired plants and their fuel suppliers were ill-prepared for the sustained cold weather. Lessons that may have been learned during previous cold-weather events over the last decade were either lost at some existing plants or not adopted at new ones.

Several industry presentations over the last six months focused on cold-weather prep and ops, and for good reason, changes are afoot. NERC has adopted a new reliability standard, EOP-011-2, which includes mandatory cold-weather prep plans, with specificity, and training to be provided to O&M staff (see accompanying article).

Some of the horror stories depicted during these user-group presentations included a plant in Nevada designed for 15F heat trace which experienced sustained ambient temperatures of -7F. Once the plant tripped because of a frozen transmitter, the rest of the plant froze up; it took a week, reportedly, to restart. Other horror stories are the plant that spent $100,000 on more robust transmitters and another which spent and additional $4.5-million to run in cold weather.

Woodbridge Energy Center has won CCJ Best Practices and Best of the Best Awards for the quality time staff invested in bolstering its heat tracing and cold-weather prep and ops; the word “inadequate” is probably charitable for describing the original heat-trace systems and other equipment the owner/operator found lacking during pre-commissioning and early operation. Get some of the details by reading the sidebar in the accompanying article, “Grid meltdown forces cold-weather prep to front burner,” based on information shared by users during a roundtable presented at the Combined Cycle Users Group (CCUG) 2021 virtual conference.

A manager at a plant CCJ visited in Kentucky said “cold weather is a pain in the ass.” Apparently, the design cold-weather points (wind, temperature) are for short duration. If they see 5F to 15F for multiple days, they have problems with insulation and heat tracing. Like Woodbridge, the plant had warranty issues with the EPC.

“We need a building!” the guy in Kentucky said. Woodbridge, in northern New Jersey and near the coast also was designed as an outdoor plant. A plant in Kansas, featured during the CCUG roundtable was described as “primarily an outdoor facility.” Most of the plants in Texas affected by the 2021 freeze undoubtedly are of outdoor design as well.

The questions nobody asked or volunteered in their presentations at these meetings are: (1) Why are plants in these locations being designed as outdoor facilities in the first place; (2) why are the ambient design conditions selected of such short duration, or not reflecting possible ambient conditions of temperature, wind speed, direction, and duration; and (3) why are these critical heat-trace systems being installed by local electrical contractors with little or no experience in this area?

Surely, someone with some financial smarts can see that the risk of one or two penalties from the ISO for non-performance during cold weather would more than pay for more robust systems. And surely many of us can speculate that the cost of complying with new NERC standards, which in itself doesn’t necessarily guarantee better cold-weather ops, could have paid for some gold-plated design margins of the kind electric utilities used to apply routinely to avoid these crises.

Revenue lost during cold-induced downtime could probably pay for a plant designed to run in Antarctica, regardless of what arguments the EPCs, banks, and project developers put up about additional capital expense, change orders, etc.

Granted, paying more for more robust design won’t solve all the cold-weather problems. It’s still easy for operators to neglect proper inspection and care upstream of winter, or these days simply not have the bandwidth with so few people on site. But better design surely is the place to start.

Utilities used to design for reliability. Today, the industry is designing for resiliency. Someone expressed the difference this way: Reliability is avoiding a punch, resiliency is the ability to take a punch and bounce back. What happened in Ercot is neither. But maybe a few hundred deaths from a statewide grid failure will cause us as an industry to go back to avoiding a punch with better design principles and margins, rather than wrestling with NERC over the wording in standards, new forms to fill out, guidelines to follow, and penalties to pay.