A flare plays a crucial safety role at many process plants. Indeed, it often provides the last line of defense against a serious incident by burning off flammable gases released. No practical alternatives to a flare (or flare stack as it’s often called) exist.
A flare primarily serves to handle gases released by pressure relief valves and other devices during emergency or equipment over-pressurization events. For instance, interruption of the usual operation of a plant, such as by failure of key equipment or a power outage, may lead to potentially dangerous accumulation of gases; sending these to a flare and igniting them via a pilot light ensures their safe combustion, thus preventing their escape into the atmosphere.
A flare also often combusts gases for relatively short periods during startups and shutdowns, e.g., to allow proper sequencing of events (such as reintroducing fluids during startups and emptying process equipment and lines during shutdowns). Many plants resort to flaring to deal with gases generated during transients in regular operations; such avoidable gas flaring should be kept to the absolute minimum possible.
Flares are a major source of air pollution such as carbon dioxide emissions. An improperly operated flare may emit hydrocarbons (methane, etc.) and other harmful gases or volatile organic compounds.
A typical flare package is simply a set of equipment that safely combusts waste gases at a pressure drop that doesn’t compromise plant relief systems or can’t be utilized. However, it is far more complicated than it seems. So, here, we’ll look at flares.
The Flare System
Typical components of a flare system include:
• a flare stack;
• a liquid seal drum or similar arrangement, e.g., a section in the upper portion of the stack, to prevent any flashback of the flame from the top of the flare stack;
• a liquid/vapor separator, usually called a knockout drum, to remove liquid from the gases;
• a pilot flame (with ignition system) that’s on continuously so it can burn relieved gases whenever needed; and
• an alternative gas-recovery system for use during partial plant startups and shutdowns as well as other times when required/possible. The recovered gas goes into a fuel gas or similar system.
Let’s now briefly turn to three aspects of a flare system.
Smoking and steam injection. Inadequate air flow will cause a smokey flare. Some smoking may occur when the flare initially lights until the flame gets enough air. In some modern flare control systems, operators use cameras to monitor the flame and adjust the operation to prevent smoking. For significant plant shutdowns such as those caused by power failures, the flare may smoke for several minutes.
Steam sometimes is injected into the flame to reduce the formation of black smoke. Using the optimum amount of steam is important because too much steam added to the flame can result in a condition known as “over steaming” that leads to reduced combustion efficiency and higher emissions.
The mixing of gases, air and water mist causes the rumbling noise frequently associated with flares. This noise increases with the rate of flaring. Loud noises might stem from extra steam to the flare and combustion of a larger amount of gas.
The gas needs air to burn correctly; this comes either via the gas flow to the flare or steam aspirators.
Flare usage and flare gas recovery. A plant can keep flaring to an absolute minimum by using waste gases instead of burning them. Flare gas recovery systems and flare gas recovery compression units, which are suitable under certain situations, have seen good progress and improvements. However, the amount, conditions and compositions of the gases delivered to a typical flare system vary considerably, creating a great challenge for any flare recovery system. Thus, unfortunately, recovering large amounts of released gases under all emergency conditions currently isn’t feasible.
Flare knockout drum. This prevents liquids from reaching the flare stack. Burning liquid droplets can spread over a large area in the plant and are a potential hazard. So, a knockout drum usually is designed to collect liquid droplets greater than a certain size, which generally is defined with respect to the details of the flare tip, particularly those related to safely burning small droplets. As a very rough indication, this size could be 200 microns.
A flare knockout drum typically is a large horizontal or vertical vessel. It generally should provide at least 30 minutes of liquid holdup from all safety/relief valves and other emergency releases. The drum also should be able to handle liquid slugs that usually occur at the start of a significant flaring event.
The sizing and design of a flare requires consideration of the full range of flaring duties in different operating and emergency scenarios, as listed in a contingency table or the like. This is a major exercise for a chemical processing plant. The selected configuration also must work within the smokeless flaring operating range.
The first set of scenarios is flaring during a whole plant emergency shutdown. This may involve a very large flow of gases that should be destroyed, with safety the primary consideration. That flow determines the necessary hydraulic capacity — i.e., the maximum waste-gas flow the flare system can handle. The second set of scenarios is for treatment of waste gases generated during normal operation, including planned decommissioning of machines and equipment in different units. While safety still is imperative for such scenarios, emissions also are important. The actual waste-gas flow rate and composition may vary significantly during normal operation but the flare still should be capable of safely destroying the waste gases while minimizing emissions.
Four performance parameters are important for most flares.
The first is the so-called smokeless capacity. This is the maximum flow of waste gases that can be sent to the flare without producing significant levels of smoke. Smokeless capacity typically at least should equal the maximum waste-gas flow rate expected during normal operation.
The second performance parameter is the thermal radiation generated by the flare as a function of the waste-gas flow rates and compositions. Radiation levels at ground level usually are limited to avoid disturbing personnel and damaging equipment. After choosing the most-remote practical flare location, stack height is set so acceptable radiation levels aren’t exceeded at ground level.
The third parameter is noise. Excessive noise can create problems for plant personnel, the environment and the local community.
The fourth key parameter is emissions produced. Flaring gases creates emissions such as nitrogen oxides (NOx), sulfur oxides, greenhouse gases and volatile organic compounds. These emissions, in combination with any unburned gases, contribute to total facility emissions. Decades ago, flare emissions weren’t specifically parameters of interest. One reason was because they were difficult to measure. However, this isn’t the case anymore. Today, flare emission reduction garners great interest. Indeed, it’s a serious requirement for modern flaring systems.
A flare must contend with environmental and plant conditions such as strong winds, storms, extremely high gas flows, etc. So, a variety of issues can arise.
Traditional flares have suffered many problems, including environmental pollution, operational difficulties, safety issues and integrity concerns. Two significant potential issues are “flaming rain” (consisting of unburned droplets of fluids) and smoke. In addition, strong crosswinds sometimes have extinguished flare flames; this can produce a high level of environmental pollution as well as operational and safety concerns.
Aerodynamics plays a major role in the chemical reactions and, therefore, pollutant formation in the flame. In general, high temperatures in the central zones of the flame lead to more pollution, particularly increased NOx production. So, avoiding very high temperatures in the center zone of the flame is important.
High winds can cause issues and disturbances to flares. For instance, in some old-fashioned flares as the crosswinds increase from below 10 km/h to above 20 km/h, the flame becomes unstable. Swirl technology can provide a solution. For example, in a modern flare design, as wind speed rises, the swirl produced by the flare also goes up and the area of recirculation gets stronger. This can stabilize the flame more and act as a pilot light to the flame. One possible arrangement uses the wind itself to cause swirling; when the wind speeds up, the inlet air speed increases, leading to higher swirl and a stronger recirculation zone. The strong recirculation area stabilizes the flame and better combustion takes place. Proper atomization of liquid droplets that passed through the knockout drum is another important factor in flare performance.
The flare tip — whose job is to produce a vertical flame standing above the flare stack at high gas flow rates — should work under harsh thermal and corrosive conditions. So, its service life is important. However, operational conditions may vary considerably; at low gas flow rates, especially with a strong prevailing wind, external burning with flame impingement on the outer surface of the flare tip often occurs. A flare tip needs very careful design. The tip nearly always is fabricated from a corrosion- and heat-resistant alloy steel. Today, alloys with higher nickel and chromium content such as 800 series, alloy 625, etc., have supplanted traditional type-300 and other stainless steels. These changes have provided some improvements but unpredictable flare tip failures still happen. Failure may be catastrophic and can occur at apparently random operating times, requiring not just tip replacement but also an unscheduled shutdown. While the cost of tip replacement itself may be relatively small, an unscheduled shutdown could result in a substantial economic penalty. For instance, a large chemical processing plant incurred an estimated $11-million hit from two weeks of unscheduled shutdown caused by a flare tip failure.
Thermal fatigue, corrosion, stress-assisted oxidation and creep typically are responsible for such failures — with thermal fatigue, the repeated stresses exerted on the flare tip during heating and cooling, thought to be the main culprit. In the presence of wind and at low gas flow rates, flame impingement on one side of the flare tip can create large temperature gradients, causing considerable thermal stresses. Sophisticated thermal studies and modern thermal imaging can quantify the levels of temperatures and stresses formed as a result of different operating and malfunctioning modes.
A Case Of Cold Gas
Let’s briefly examine why a process plant ran into the serious issue of cold gas in the flare system. As part of a renovation/expansion program, the facility added processing units; it also put in a new flare stack because the old flare was small in size and capacity. At the final review before the commissioning, it was found that the minimum operating temperature of the flare system (flare piping, flare stack, etc.) was only -15°C. (Apparently, details were copied and pasted from the old flare.) However, this wasn’t suitable because the gas pressure in new units, particularly the high pressure compressors, exceeded 100 Barg. In case of blowdown, the gas could reach low temperatures, below -15°C. The travel between the high pressure discharge of the compressors and flare stack would raise the gas temperature somewhat. However, the volume of low temperature gas in a full blowdown could affect the flare system. Based on accurate simulations performed after this finding, at the highest possible pressure (relief pressure of the high pressure compressor unit) and ambient temperature of about 2°C (winter), the lowest gas temperature at blowdown could be around -47°C. Yet, the provided piping and flare only could stand a temperature of about -29°C.
This led to ordering a new flare stack and flare piping with proper materials and low temperature capability (below -47°C) to replace with existing flare system. However, delivery time was four months.
Since it was summer, rather than wait, it was decided to commission the plant but impose some restrictions. Operating procedures were developed for the four-month interim period to maintain gas temperature at 25°C or more at all times, to keep the minimum operating temperature in the flare system above -29°C. Also, a site standing instruction mandated avoiding blowdown or de-pressurizing after prolonged trip or non-vented shutdown where gas temperature (static inventory) might cool down below 25°C.
Then, when the new flare stack, piping, header, etc., were delivered (within four months), the flare system and piping were replaced in a short one-week plant shutdown.
An Emissions Example
Let’s now look at another case study to compare the performance and emissions of a traditional flare with a modern one. In the old-fashioned flare, the flame was long, yellow and highly luminous, indicating poor mixing. Within a distance of around two-to-three times the flare stack diameter, the flame was a simple diffusion one. Considering a zone of six times the flare stack diameter, the flame temperature went from 1,400°C at the center to 1,000°C in the middle and 800°C at the boundary. The central high-temperature zone created considerable pollution, particularly a high rate of NOx production. Lower temperatures at the boundary were due to more entrainment of air there. Combustion efficiency was estimated at less than 90%.
After replacement with a modern flare, the temperature of the central part dropped to about 1,250°C and zones of temperatures above 1,000°C were substantially reduced. Large portions of the central section previously in the 1,400°C zone were converted to areas of 1,000°C or lower. The modern design and improved aerodynamics provided combustion efficiency estimated at more than 94%. Moreover, due to better temperature distribution, NOx production fell to less than 30% of that of the old flare.