Fenix Hydrodynamic Steam Traps
A hydrodynamic steam trap (or venturi trap) is a steam trap that works by forming a condensate steal to prevent the loss of steam through the trap. Conventional mechanical traps (float, thermodynamic, thermostatic, inverted bucket, free float) all have a part/s that move. With the hydrodynamic trap there are no moving parts only the water (condensate) seal that adapts to the changing processes – hence the name hydrodynamic steam trap.
The trap has a carefully sized nozzle and venturi. The condensate continuously arrives at the trap and forms a seal (basically flooding the nozzle). The steam pushes the condensate through the nozzle but more condensate arrives to maintain the seal. So long as the condensate passes through the nozzle at the rate it forms then it will not be possible for the steam to pass through the hydraulic seal created by the condensate. The discharge of the nozzle is a staged venturi. The discharge of the traps is always at a lower pressure and therefore some of the condensate turns to flash. By mass the flash may be only 10% but by volume it is hundreds of times greater than the condensate. The sudden expansion is like a rocket and creates a counter pressure that controls the flow of condensate through the nozzle.
Yes, and very well. The flash steam that creates the back pressure, which in turn regulates the flow through the trap, varies with varying flow. With lower flow and pressure there is less flash and vice versa. So as the process conditions change the capacity of the trap self regulates. The result is a trap with no moving parts that can operate on maximum load but turn down to minimum load.
Yes and no. If you take a typical process application such as an air heater or fluid heat exchanger then the trap will work over the full range of that application. For example if the maximum steam flow of a unit is 5t/hr at 150 psi (10 Bar) and 1.5 t/hr at 45 psi (3 Bar) then the trap will work fine over the full range, down to the point where the control valve fully closes.
However you cannot take the same trap that is designed for 5 t/h-150 psi and use it on another application that is design for 2 t/h-150psi. But this is the same as control valves. Each has to be specified for the application. So every process trap is individually sized for each application.
In the both line drainage applications and trace heating the pressure is nearly always constant however the load is very constant too. With line drainage the variation in load is just a function of the outside air temperature change but even in extremely varying climates the change relative to the steam temperature is small and with the correct insulation the affect on the amount of condensate produced is minimal. When it does become cold there is a slight sub cooling of the condensate before the trap and this reduces the amount of flash produced and therefore increases the capacity of the trap. Again, a self regulating function.
In the case of trace heating the principal is the same but in many ways more flexible because of the section of line before trap which can dissipate a lot of heat if condensate starts to build up. Hence the hydrodynamic traps self regulate to compensate for changing loads in the trace heating system.
Many. Normally we focus on the energy saving because this is a tangible benefit that can be measured relatively easily. Typically if a process plant is retrofitted with Fenix Traps the savings will be between 10 and 25% of the current steam use (for trapped steam, excluding direct injected or turbine use). This saving alone is sufficient to justify the investment in the Fenix Traps and normally provides a payback of less than 12 months (including installation costs).
Another benefit is the reduction in maintenance costs. Once Fenix Traps are installed it will not be necessary to change the traps again and they will last the lifetime of the plant. The Fenix Traps are guaranteed for 15 years but in reality are never likely to fail. Provided there is only condensate (water) passing through the nozzle then there is no erosion and therefore the traps will not fail.
However the real benefit is the reliability of the plant. A typical survey of a process plant will reveal traps in various states of repair: failed open, failed closed, rapid cycling, flooded and typically less than 50% ok. But the traps have a function so it’s not only the trap itself that is not working well but the associated equipment. For example, if a trap on a U tube heat exchanger has failed open or is rapid cycling then as well as wasting steam there will also be added erosion of the tube bends due to the steam passing through the tubes at a higher velocity. If the loss through the trap is very high then it can result in the heat exchanger not being able to maintain pressure and therefore will have reduced output.
So a complete conversion to hydrodynamic traps that don’t fail has an enormous affect on the overall reliability and performance of a process plant.
Fenix (or Fenix associated companies) are responsible for the sizing of the traps. For process applications this is similar to a control valve i.e. flow and differential pressure and each trap has to be fitted to the specified application.
For line drainage and trace heating applications the traps can be supplied with nozzles of different capacities: D -Dry/superheated steam, L- Low load, M- Medium load, H- High load, V- Very high load, E- Extremely high load and N – Non Standard (to be sized). Most applications would fall between D and V.
For Line Drainage and Trace Heating applications the traps can be stocked on site and the correct nozzle fitted during installation.
With both the process trap and the line drainage/trace heating traps it is easy to change the nozzle to change the trap capacity if the original is found not to be correct.
Obviously the costs changes depending upon the model and size but in comparison with mechanical traps Fenix Traps are typically twice the price. However even taking conservative values for steam losses and failure rates of mechanical traps, the Fenix Traps turn out to be between 2.5 and 3 times better value for money – and this doesn’t include the enormous improvement in the performance and reliability of the process plant.
All, except traps used for overflow applications or on the bottom of flash vessels (where we prefer to use a control valve). The Fl trap is a flanged trap that can be supplied with various flange ratings and sizes up to 4” (100mm) – for process applications (can be designed to have the same face to face dimensions as the trap being replaced). The IVP is for lower flow process applications (maximum 1 ½”, 40mm) and can be socket welded in line. AM is for process applications at lower pressures with screwed connections. The EF trap has similar dimensions to a traditional thermodynamic trap and can be either screwed or socket welded. The QF trap is a universal two bolt design that fits on either a Fenix adapter or any of the competitors units.
When compared with traditional mechanical traps there are no worthwhile comparisons, they are light years ahead. There are other companies in the market selling hydrodynamic traps but Fenix Earth has been working with this product for many years and has designed a product range that not only works but is also easy to install. This has come about from experience that shows that if there is too much effort required to change from one type of trap to another it is less likely to happen. By designing the product range to have very little installation demands it makes the whole process of plant conversions more probable.
A Condex is a sophisticated heat exchanger that is designed to capture both sensible heat and latent heat from exhaust gas flows. The most typical application is the exhaust gas flow from boiler flues but can be used for CCGT exhaust, ovens, or dryers.
Traditional boiler economizers are designed to heat the water from a de-aerator or condensate tank before entering the boiler. The general rule is that the exhaust gases should not be reduced below 230oF (110oC) to prevent localized condensing of the exhaust gases. The resultant acidic condensate can cause erosion problems in the chimney. The Condex is designed to operate below the dew point of the gases (normally around 160oF, 70oC) and is designed to operate with the acidic condensate that is formed.
The simplest application is to heat the softened boiler make up water before it enters the de-aerator. Typically softened water will enter the de-aerator around 70oF (20oC) and needs to be heated to the de-aerator temperature, 220oF (105oC) with live steam. The Condex can preheat this softened water up to 195oF (90oC) before it enters the de-aerator and therefore significantly reducing the quantity of live steam required in the de-aerator.
However if there is no requirement for heating softened water then it is possible to use the Condex to preheat process water and displace live steam required for process heating.
No. They are designed to be completely redundant and can be switched on and off without affecting the boiler. Not only that but when they are up and running the units self adjust with the modulations in the boiler output. They utilize a system which dilutes the exhaust gases very slightly with air drawn down the chimney stack. The flue gases cannot pass the flow of air coming down the chimney so it guarantees that all of the flue gases are pulled into the Condex unit. A damper on the duct to the Condex modulates to control this process and as the boiler load changes the dampers adjust accordingly to always maintain this slight dilution. The result is a system that is fully self regulating and the operator does not have to adjust any parameters.
The other great benefit of this system is that several boilers can be connected to one Condex unit. Each boiler would have its own damper to control the exhaust flow and therefore several boilers can be operated fully automatically with no operator input required.
The big advantage of the Condex unit is that it is designed to operate below the dew point of the flue gases. This means that the majority of the latent heat in the water (byproduct of combustion) can be recovered. The water (which is in gaseous form) condenses in the Condex and releases the latent heat. The resultant condensate that is formed is acidic due to the contact of the water with CO2 but the unit is designed to handle this acidic condensate. The construction of the heat exchanger is stainless steel and aluminum and the base and chimney of unit is fiberglass.
The condensate that is formed is not strongly acidic and so in many cases can go direct to drain. Alternatively the condensate can be treated with neutralizing salts and re-used in the system.
Heavy fuel oil (No.6) invariably contains some sulphur. When the condensate forms in the Condex the sulphur dissolves in the water to produce sulphuric acid which is very corrosive. However it is possible to use a special coating of the Condex heat exchanger tubes to reduce the affects of the corrosive condensate. The fiberglass base and chimney of the Condex unit are not affected by the acidity.
So, although the preference is to use the Condex on natural gas applications, it can be used with Heavy Fuel Oil if the additional coating is applied.
There are many advantages to the Condex unit (which is indirect condensing) compared with direct contact units, the main one is that Condex can recover the sensible heat in the exhaust gases as well as the latent component. With direct contact the maximum temperature is limited by the dew point temperature of the gases which for a boiler will be less than 160oF (70oC). With the Condex unit the water can be heated to 195oF (90oC) and therefore recover more energy.
The other big disadvantage of direct contact is that the water to be heated comes into direct contact with the flue gases and becomes acidic and needs to be treated. With the Condex there is not contact between the water being heated and the flue gases.
With the Condex unit it is very simple to measure the savings. A flow meter on the water supply together with temperature transmitters on the water entering and leaving enables the total energy recovered to be calculated.
For industrial processes that can be supplied from one small steam boiler there is an inline Condex unit available. This cylindrical design fits directly in existing boiler chimney and doesn’t require the fan/damper system of the full scale Condex unit. However it is only intended for single boiler use on smaller scale boilers.
Usually the return on investment is less than 3 years including the cost for the intallation.
Paper Machine Steam And Condensate System
Our experience is that most steam and condensate systems are designed over simplistic. For a competitive paper machine it’s analogous to having a Formula One car with state of the art aerodynamics and suspension but with an old V8 with carburetors. The steam system is actually a combination of a Temperature Balance, Energy Balance, Mass Balance and Pressure Balance. All four of these must be working in harmony not to have any problems. Added to this, paper is a natural product and has to be heated in a controlled manner to prevent quality problems. And finally the systems need to be efficient.
The objective of a steam and condensate system is to remove the condensate from the cylinders at the rate it is being formed. But the condensate has to be removed evenly to prevent hot/cold spots and consequent profile problems, and in a way that doesn’t cause mechanical problems (erosion) or poor system efficiency.
To achieve all of this, especially on machines that make a wide range of products, is not so easy.
First the sheet has to be heated in way that does not damage the surface of the sheet (picking) so the first section of dryers has to correctly specified to control the sheet warm up. If there is a unirun then this has to be considered in the steam system design. Cylinders should be grouped to have similar condensing loads. If there is a cascade then the ratio of the number of dryers in each group has to be correct to utilize the blow through steam. As well as having the correct mass of blow through steam it is important to have the correct velocity, too low and the condensate will not be picked up, too high and there will be excessive differential and erosion. The siphon diameter has to be correctly selected to match both the blow through steam mass flow and velocity and will vary depending upon the position in the group. As the machine speed increases then the size of the steam groups should decrease. At the dry end of the machine the condensing load drops off rapidly so a smaller steam group should be used, more so on fine paper machines with lower moisture. With size press applications the sheet after the size press has to be heated carefully. And the list goes on. So the design of a paper machine steam and condensate system is not so simple.
Velocity Flow Control (VFC) controls the velocity of the blow through steam in the siphon riser. Assuming that the siphon has been correctly sized to give the correct amount of blow through steam then by controlling the velocity of the blow through steam in the siphon it will always ensure optimum condensate removal. It is not possible to measure the blow through steam directly in each siphon in the dryers but the total blow through steam for each group is measured and the velocity calculated for each siphon (assumes the condensing load is the same for each cylinder in that group).
The velocity of the blow through steam is controlled by modulating the differential across the group. An increase in differential will increase the velocity and vice versa. If a cylinder begins to waterlog there will be a reduction in the siphon velocity and the control loop will increase the differential to clear the condensate. If there is a sheet break and very little condensate then the differential will be reduced to prevent an excess of velocity in the siphon. So velocity control is actually a system for optimizing the differential for all operating conditions.
Spoiler bars are installed to improve both the heat transfer from the steam to the sheet and to obtain an even heat transfer across the width of the dryer. When the condensate starts to rim (around 1000 ft/min, 300 m/min for 5ft dryers) there is a still a lot of turbulence in the condensate rim due to slippage. As the speed increases further the slippage decreases and there is less turbulence until the condensate rim is almost static. Condensate is a very poor conductor of heat so relies on the turbulence to transfer the heat from the steam to the cylinder wall. The turbulence bars are employed to break up the condensate rim and reintroduce the turbulence. At speeds of below 2000ft/min (600 m/min) there is very little benefit because of the natural turbulence in the condensate rim. As a guide the bars should be considered for speeds of 2200 ft/min (650 m/min) and above.
However if high speed stationary siphons are used then they bars should be fitted irrespective of the speed to ensure a more even cross machine temperature profile.
First we have to define the type of stationary siphon. There are two broad categories, low speed and high speed. Low speed stationary siphons are often just a bent tube or a tube with a flexible knuckle joint. High speed stationary siphons are a very sophisticated piece of engineering equipment with a cantilevered support system. Low speeds stationary siphons are only suitable up to around 600 ft/min (200 ft/min). Above this speed the choice is between high speed stationary siphons or rotary siphons.
Before making a decision, if the cylinder is old and has a journal diameter of less than 3 ½” (90mm) then it becomes difficult to fit the high speed stationary siphons. The big advantage of the high speed stationary siphon is the required differential remains low even at high speeds. So above speeds of 3000 ft/min, (1000 m/min) then high speed stationary siphons are a must. Between 600 and 2500 m/min there is very little to choose between the two options if the steam system is well designed but as the machine speed increases the stationary siphons are more favorable.
Stationary siphons are more flexible requiring lower blow through steam flows and low differentials and therefore can work better on a poorly designed steam and condensate system. They can work better on differential control since the differential required does not increase with machine speed.
The big disadvantage of the high speed stationary siphon is the cost since they usually require spoiler bars too and come complete with an integrated joint. Once fitted they usually run trouble free. However our recommendation is still to control the blow through steam flow with velocity control and vary the siphon diameter depending upon the condensate load. Often a cheaper option is used which has a fixed riser diameter for all dryers but uses orifice plates in the condensate drop to control the blow through steam. Two phase flow erodes orifice plates very rapidly so we don’t recommend this.
If the machine is already fitted with suitable sized rotary joints and is operation below 2500 ft/min then we would stick with rotary siphons. Above this would depend on the characteristics of each machine and the economics.
Both have their merits. The advantage of the thermocompressor system is that it allows the steam groups to operate independently. However there are often natural limitations to steam group pressures and in these cases a cascade system works perfectly well. An example of this is in the wet end of the steam groups where the sheet has to be warmed carefully and therefore the steam groups have a natural cascade where the steam pressure increases in the machine direction. The other example is when there are two different types of dryers on the same machine with different maximum operating pressure. Again there is a natural cascade with little benefit to using thermocompressors.
Thermocompressors are useful on larger machines or low pressure machines where it is difficult to keep increasing the pressure in the subsequent group. A large machine with a cascade system can struggle to make lighter grades and there is also a potential loss of capacity at maximum production. In this case the thermocompressor system is a good option. The main disadvantages of the thermocompressors are the additional capital costs (of the thermocompressor and associated safety valve) and the need to have a high pressure steam supply. As a rule of thumb the motive pressure for the thermocompressor needs to be twice the maximum operating pressure. Usually the high pressure steam is available but may need a new line from the boilerhouse. If there is a steam generator then the preference is to use as much low pressure steam as possible so taking off a supply for thermocompressors can affect the electricity generation.
So each machine has to be considered individually and the right system, which is often a mix of both cascade and thermocompressor systems, has to be selected depending upon various parameters.
This is a frequently asked question that is difficult to answer. One of the biggest factors affecting the steam production is the press section. The ex-press moisture can vary from 60% down to less than 50%. The affect of the press moisture on the steam consumption is enormous. With a high ex-press moisture not only is there more water to remove but there is also more water to heat up to the evaporation temperature. Another important factor is if there is a size press since this is affectively two drying machines.
It is important to distinguish between drying effectiveness and efficiency. A system can be efficient but not effective and vice versa. One example is where condensate is cascaded to lower pressure condensate tanks. This is efficient because it utilizes some of the sensible heat in the condensate but not necessarily a good idea because the flash steam can have an adverse affect on the control of the blow through steam and have a negative effect on the product.
One of the key issues is to ensure that the system is not venting steam to atmosphere which maybe a thermocompressor design issue, incorrect configuration of cascade groups or siphon design. But assuming that there is no steam venting then there are still many design factors that affect system efficiency. It is important to consider both the energy for the steam system and the energy for the hood. Ideally a machine would have a fully enclosed hood with pocket ventilation. With a good pocket ventilation system, the action of the air will reduce the sheet temperature and increase the overall heat transfer from the dryers and increase the maximum drying capacity. But the pocket ventilation air needs to be heated and the best way is to use the sensible heat from the group condensate tanks and flash steam.
With a fully enclosed hood there is often enough energy in the condensate and flash to meet all the pocket ventilation heating requirements. So when considering the system efficiency it is important to consider the total energy requirement for the dryers and the hood.
A poor steam and condensate system design will adversely affect many parameters such as sheet moisture profile, high drive loads, erosion of system pipework and ancillaries, sheet breaks, poor system flexibility, increased operator input and product quality. A well designed system will not only perform well in the moment but will continue to perform well for many years. For example, if the siphons are well matched to the condensing load then there will good condensate removal but also very little erosion so the system will keep running ‘as new’.
If the system is always in balance then there is less probability of water logging which can result in sheet breaks. With fewer sheet breaks the system is more stable. With a well designed system and good controls it is not necessary for the operators to make adjustments to keep the system working only dial in the required pressures for the production requirement and even this can be automated.
PES has many years specializing in designing steam and condensate systems for dryer systems. The PES focus has always been to provide the system available in terms of production performance and efficiency. Although PES does manufacture a rotary siphon the core business is system design rather than selling product. As such we will offer what we believe to be the best possible system and try where possible to utilize as much of the existing system.
Our key philosophy is to offer only what is needed. Sometimes this is a complete new system with new hood. PES works with a Canadian hood supplier, Enerquin Air, who we consider have a similar philosophy to PES and supply the highest quality product. Other times we can just make some minor changes and that is all that is required. When we look at a machine we ask ourselves ‘if this is my machine what would I need for it?’, as opposed to just selling as much equipment as possible.
With over 25 years of experience designing steam and condensate systems PES has some justification in claiming to be the best available.
HRSG – Heat Recovery Steam Generator (pronounced HERSIG) – is a unit used to generate steam from waste heat air streams. The most typical use for a HRSG is generating steam from the discharge gases of a gas turbine. However PES specialize in HRSG for recovering waste heat from process applications, the most common being Yankee tissue machine hoods but can be applied to other applications such as ovens or other dryer applications.
The key issue is the temperature. The gases need to be considerably hotter than the steam generated saturation temperature. Provided there is around 160oF (100oC) difference then there is a possibility to generate steam. The next fact is the flow rate, more flow more steam and to a lesser extent the humidity.
In many ways they are similar and the designs are based on conventional boilers. The HRSG are designed according to ASME codes used for boiler design. However there are differences the first being there is no burner. Because the gas temperatures are much lower than the flame temperatures the surface area required to generate a quantity of steam are considerably higher for HRSG than boilers. The controls for the HRSG are similar to a boiler, level control, level switches, pressure switches etc.
The HRSG are usually designed to be redundant, that is they can be switched on and off without having any affect on the process. This is important because they are an energy saving device and core to the production process. The HRSG are supplied with dampers that allow them to be bypassed. It is important to have several interlocks with the pressure and level and the dampers.
The simplest system is to generate high pressure steam that feeds into the thermocompressor, typically between 235 and 250 psi (16 to 17 Bar). There are very few controls required since the HRSG will never supply all of the steam required and therefore the high pressure line to the thermocompressor will dictate the boiler pressure.
The disadvantage of this is that the steam is relatively high temperature, 400oF (205oC) and therefore the gases will the HRSG at some temperature higher than this, still with a lot of energy. The option to generate steam at the Yankee pressure around 100psi (7Bar) is a bad idea because there will be more steam generated than passes through the pressure control valve. The pressure control valve will close completely and then start to back off the thermocompressor which will affect the condensate removal from the Yankee. Not a good option.
The final option is to generate low pressure steam 15psi (1 Bar) but this is only an option if there is sufficient and continuously demand for the low pressure steam. Generating low pressure steam usually requires a better understanding of the site energy use but typical applications that can be considered are machine steam box, boiler house de-aerator, stock prep, space heating or process water heating.
Because of the lower saturation temperature of the steam at 15 psi (150oF, 120oC) it is possible to extract much more heat from the exhaust air stream and therefore generate more steam. Increase can be as high as 60% compared with generating high pressure steam. However it is necessary to fully understand the steam use to be successfully implement a low pressure option.
If there is an existing air to air heat exchanger to preheat combustion air and make up then there will be a reduction in the off coil temperatures and therefore will require some additional gas use for the hood to compensate. In the majority of cases this is not a problem because the exhaust air flow rate is considerably more than the combustion and make up air and so much more energy will be recovered than the extra amount added. But it is necessary to consider this additional gas demand when doing the economic analysis to justify the project investment.
The only case when a HRSG maybe not be viable is if the steam for the plant/Yankee is being supplied by a CHP plant which sometimes have a minimum load demand and if the load is close or below this value then saving steam is not an option.
There are two broad types, fire tube and water tube and there are pros and cons to both. The fire tube (or shell and tube) is a simple design with doors either end that will allow access to clean the tubes. The advantage of the fire tube design is that it can be delivered almost ‘plug and play’ with all instrumentation, controls, dampers, valves etc all ready preassembled. Once in position it is then just a question of connecting the steam, condensate, electricity and compressed air. The disadvantage of the fire tube design is that they are heavy when filled with water (typically between 30 and 40 tons). If the mezzanine floor cannot support this weight then they have to be located on the ground floor level which often requires longer duct runs which increases installation costs.
The water tube type can be made more compact and lighter. The PES design is a single dome design and this can be located separately to the heat exchanger section to minimize the loading on the mezzanine. The disadvantage of this design is that it requires some assembly on site.
There is always some fibers that are carried in the exhaust of the hood however if the exhaust temperature is maintained above 500oF (250oC) then the fibers will burn. There will still be some ashes from the burnt fiber but this is in the form of a dust and gets carried through the HRSG without sticking to the tubes.
Of the units we have operating there doesn’t appear to be a need to clean the tubes. There is a light coating of the dust on the tubes surface but no build up. We would always recommend inspections every 6 months to check.
In the case of the fire tube design the tubes can be brushed out with normal boiler tube brushes. With the water tube design we recommend cleaning with a power hose.
In general the HRSG are low maintenance items. The most probable items that require maintenance are the multistage condensate pumps. With a correctly design system and correctly specified pump the maintenance should not be excessive but this type of pump usually requires some periodic maintenance.
Other maintenance are just the usual items such as control valves, transmitters etc. It maybe that no maintenance is required but it is always worth making a routine check on all items.
The TDS level control is exactly the same as for a conventional boiler. As well as maintaining the correct TDS level it is necessary to purge the bottom of the HRSG once a day, as required for conventional boilers. Because the evaporation rate is relatively low compared with the volume of water and because the condensate comes from condensed steam from the Yankee, the increase in TDS in the HRSG is usually very small. It has been found that the daily bottom purge is sufficient to maintain the correct TDS level in the HRSG but each application will be different and it should be the responsibility of the boiler hose manager to maintain the HRSG in the correct state.
The saving from installing a HRSG on a Yankee tissue machine will come from reduced gas to the steam boilers and the payback will be a function of the gas price. It will also depend on which steam pressure is selected with a lower pressure giving a faster payback.
Typically a HRSG project for a tissue machine will provide a payback of less than 3 years including the installation costs.