I am a newcomer to drum level technology. Educate me!

I understand drum level technology, but I need to understand the error problem...

I understand the error problem now. What can be done to solve the problem?

There are several different new technology instruments. Which one works best?

Prove it. How do I know that these instruments really tell the truth?

What does ASME Boiler Code require me to do?

What does R-Levels recommend for my boiler/ HRSG? Explain!

How do I get in touch with R-Levels?

 

I am a newcomer to drum level technology. Educate me!

Let's start with a simple example. If we have a simple water tank, we can install a glass water level gauge that connects the bottom of the gauge to the bottom of the tank, so gravity makes the inside and outside of the tank levels equalize, so we can see exactly what the water level is inside the tank. Now let's consider a more complicated example. Let's say that the water inside the tank is under pressure. A simple glass gauge connected to the bottom of the tank (and open to atmosphere on top) will turn into a fountain of spraying water! To overcome this, we build a gauge that is not open to atmosphere. It is instead connected to both the top and to the bottom of the tank so that the static pressure is equalized. Now gravity does its thing and we can see exactly what the water level is inside the tank by observing the sight glass outside the tank. Problem solved. Now let us make the example a little bit more complicated. We fill the tank with hot, pressurized water. If you are familiar with thermodynamics, then you already know that water density changes - a lot - depending on temperature and pressure. For purposes of this discussion, let's say that the hot water inside our tank has a density 2/3 of the density of cold water. Since the gauge glass outside the tank is cold and the water inside is hot, then if our gauge glass shows that the tank is full of 2 meters of water, then the actual level inside the tank is 3 meters. The force of gravity didn't change - the density did. As a result, our indication is 1 meter off! 

 

If you understood the previous paragraph, then you have all of the tools you need to understand boiler drum water level measurement. The only difference between the last example in the first paragraph and drum water level technology is that using a flimsy glass gauge glass sounds like a bad idea when dealing with high pressure water and steam. There are three distinct functions that we want to be able to do when we measure the water level. One, we want to be able to see the level (of course). Two, we want to be able to connect the boiler feedwater control system up so that it can control based on the level is senses. Three, we want to be able to hook up an automatic boiler protection system so that if it senses either very high or very low level, it can automatically shut down the boiler. The following three paragraphs summarize the "state of the art" with traditional drum water level systems.

 

Boiler Drum Water Level Sight Gauge

This is the technology that is almost identical to the example in the first paragraph. Various manufacturers make gauge glasses suitable for high temperature and pressure. As previously explained, this technology is susceptible to substantially large errors due to the large difference between the density of the relatively cool water inside the sight gauge and the much hotter (and lower density) water inside the drum. The best aspect of this type of gauge is that one can actually SEE what is happening inside the drum (even if some math is required to figure out the true level). The problem is that this kind of gauge cannot be hooked up to either a control system or a boiler protection system. It is good for indication only.

 

Conductivity Probe Instrument

This technology is also pretty similar to the example in the first paragraph. There is an interesting twist, though. Instead of looking at the level with our eyes, we will "look" at the level using a trick. We place conductivity probes (they look exactly like spark plugs from a car engine) at various elevations. It's fairly typical to set up one of these gauges with its conductivity probes arranged something like this: -15 inch, - 10 inch, -8, -6, -4, -2, -1, 0, 1, 2, 4, 6, 10 inches. Thus a fairly large range can be covered without using too many probes. When water touches a probe, the conductivity will change. The change is so clear that it's fairly easy to set up some electronics that can easily distinguish between water and steam. Typically, a light turns green for water and red for steam. This gauge is susceptible to a fairly large error due to the difference between the density of the water inside the instrument and inside the boiler drum, so some compensation must be done to correct for this error. This kind of instrument is not really suitable for control for two reasons. One, it has a discrete function (not continuous) output. Two, it can be set up completely independently of the DCS, so even in the event of a DCS failure, the boiler protection function can still operate. For these two reasons, this instrument is preferred for boiler protection functions.

 

DP Type Level Instrument

We need an instrument suitable for control. This is typically accomplished by using a differential pressure (DP) transmitter. We need to measure DP across something, but what? The typical solution is to place a constant head chamber on the "low side" of the DP transmitter and to feed the sensing line from the bottom of the drum to the "high side" of the DP transmitter. With this setup, we have the slightly disconcerting situation that the DP transmitter always reads a negative number, and as drum level increases, the DP decreases. We have to deal with the density difference between the "low side" and the inside of the drum. The only way to calculate the level inside the drum correctly requires us to decide three things: Assumption 1) what is the density of the steam inside the drum? Assumption 2) what is the density of the water inside the drum? And Assumption 3) what is the density of the reference leg (connected to the constant head chamber)?

 

These three assumptions are important, because the accuracy of the DP Type Level Instrument depends on getting all three right. Typically, the densities are calculated by making the following three assumptions: Assumption 1) the steam inside the drum is exactly at saturation (i.e. the boiling temperature for the pressure inside the boiler), Assumption 2) the water inside the drum is exactly at saturation and Assumption 3) the water inside the reference leg is at ambient temperature. With these three assumptions, it's relatively easy to use the ASME Steam Tables to simply calculate the densities. With this done, it's just a matter of some algebra to derive a relationship between the DP transmitter's output and water level inside the drum. All of this happens inside the DCS, so if we configure it correctly it will work correctly. With the addition of the DP Type Level Instrument to our arsenal, we now have an excellent instrument that can be hooked up to the feedwater control system. Problem solved. Note (the problem is NOT solved, so we will deal with the three assumptions in the next section). 

 

Magnetic Type Level Instrument

Now we're going to move into an area which will not be familiar to you. We asked ourselves, is there a way to develop an instrument that isn't susceptible to freezing (like the DP instrument), or susceptible to leakage (Conductivity probe instruments have a lot of penetrations), and isn't susceptible to boil-out? This led us to consider using another old idea - a high pressure float. A high pressure float will read water level very directly, but the rise of the float in water will depend on the pressure inside the drum. The float will rise higher in low pressure and float lower in high pressure, but this is easy to compensate mathematically. To deal with the "bad assumptions", we will do the same technique of steam jacketing the high pressure float chamber.

 

So far, we have a good design. The level of the float, with a steam jacketed chamber and with a little mathematical compensation, will read drum water level very accurately. But how do we read the output from the float? Our company decided to use a separate chamber. We connected a movable magnet to a rod, and connected the rod to the float. The magnet is in a part of the instrument which is NOT steam jacketed and has zero flow, so the temperature of this chamber is essentially ambient. Using a commercially available transmitter using the principle of magnetostriction, we translate the position of the float into a 4-20 ma signal.

 

This instrument, takes the best qualities of our other instruments and puts them in one place. With this new instrument, we have a technology which is not susceptible to freezing, and far less susceptible to leakage.

 

One more thing. We actually have developed two slightly different renditions of this instrument. One has an external, steam jacketed chamber, much like a conductivity probe gauge. The other design places the high pressure float directly into the drum water, so the only part of the instrument outside the drum is the magnetic component. When possible, it is preferable to use this configuration.

 

I understand drum level technology, but I need to understand the error problem...

There is a serious problem with traditional boiler drum water level measurement systems. Let us start with a story. When the author of this website was a young engineer, he had the opportunity to see the inside of a boiler drum for the first time. There was a very clear, obvious rust line on the inside of the drum that clearly indicated where the water level had been controlling for the previous year's operation. It was perhaps 9 inches higher than expected. That led to many questions, a lot of work on mathematical equations, but no clear answers. In short, it was obvious that there was a problem, but it was not obvious why. To understand why, we need to re-examine those assumptions.

 

Assumption #1. Good. The steam inside the boiler drum is exactly at saturation. Is this a good assumption? Yes, it's a very strong assumption. The drum steam is full of tiny water droplets, making it impossible for this steam to be superheated, and if the steam loses energy, it will simply collapse to water. So this assumption is good, and it stays on the list. 

Assumption #2. Bad. The water inside the boiler drum is exactly at saturation. Is this a good assumption? To answer this, let's see where water entering the boiler drum comes from. Most of the water (and steam) rises up from the steam generating (waterwall) tubes. That water is certainly right at saturation. But some of the water is coming from the feedwater system via the economizer outlet. Is this water at saturation? No, it's not. Boiler designers avoid boiling inside the economizer because that will make the boiler "burp", which would cause control problems (and probably terrify operators). So this feedwater is well below the boiling point. This mixes with the saturated water from the waterwall tubes and the resulting mix is somewhere below saturation. So this was actually a bad assumption. 

Assumption #3. Bad. The water inside the reference leg (coming from the constant head chamber) is at ambient temperature. This doesn't require detailed explanation to see the problem. It's not even possible to touch the constant head chamber, it's so hot - and it's not possible to touch the reference water leg until we are relatively far from the constant head chamber. This is a bad assumption, too.

 

So what is the result? Instead of going into the mathematics, let us simply say that Assumption #2 causes the actual drum water level to be far lower than the indicated drum water level. Assumption #3 also causes the actual drum level to be significantly lower than the indicated drum water level. Bad news! Rule of thumb, if your boiler pressure is below 1,400 psi (9.7 MPa), then the error is small enough to be less concerned. But the higher the pressure, the higher the concern. The worst situation is for subcritical boilers with steam conditions of 2,400 psi or higher.

 

Then, what is the reason that we usually discover, in an inspection, that the drum water line inside the drum is higher than the design level?

All of the drum level indicators should have the same result during the operation.  But because of the difference of the structure, installation and working ambient temperature, the instruments show a large difference among one another in the real operation condition.  But we need for the instruments to agree with one another so that we can control and protect the drum level, so we are forced to correct the drum level manually to achieve agreement.  The most common way to do the ‘correction’ is to adjust the DP transmitter output based on the level shown by the boiler gauge glass.  The gauge glass is a U-Tube theory instrument and it is usually installed in an ambient temperature that is much lower than the temperature inside the drum.  The error caused by the gauge glass makes the actual level is higher than the measured level.   To correct the DP transmitter based on the gauge glass will cause the actual level to be higher than the measured level.   When the operator controls the drum level at its normal level, the actual level is higher than the measured level and it causes the drum water inside the drum to be higher than the design level.    

 

I understand the error problem now. What can be done to solve the problem?

The underlying problem with traditional instruments are the two bad assumptions (the water inside the drum is NOT at saturation, and the water in the reference leg is NOT at ambient temperature). Now that we know about the sources of error, can we do something innovative to reduce or eliminate the problem? Let us first consider a DP Type level instrument. There is an innovative way to deal with this - instead of placing the constant head chamber and the reference leg OUTSIDE the boiler drum, place the constant head chamber INSIDE the boiler drum. This solves the problems of both of the bad assumptions in a single stroke. With the reference leg inside the drum, the water inside the reference leg is now exposed to the exact same set of steam and water conditions that actually exist inside the drum. Now there is no mathematical correction that needs to take place. To be completely clear, now the reference leg is not assumed to be ambient so that error has been eliminated. Furthermore, the part of the reference leg above the water level is immersed in saturated steam so it will exactly match that temperature. The part of the reference leg below the water level is immersed in slightly under-saturated water, so it will exactly match that temperature. The result is zero mismatches between the reference leg density and the densities of the water and steam inside the drum. As long as the portion of each sensing line that goes outside the drum are of matching length and routing, this kind of instrument will read exactly right, using a very simplified mathematical relationship in the DCS. Problem solved!

 

What about the conductivity probe and the boiler drum water level gauge glass? Once again, we understand the sources of error, so is there an innovative way that we can reduce or eliminate the errors? These instruments need to be outside the drum, so we can't employ the same trick we used on the DP Type instrument. So let's try a different approach. Let us match the temperatures inside and outside the drum. We can accomplish this by "steam jacketing" the instrument. By putting drum steam all around these instruments, we can get the water inside the instruments to virtually exactly match the temperature inside the boiler drum. If this works "as advertised" then a water gauge glass level should exactly match the actual level inside the boiler drum with no mathematical compensation required. The same thing goes for a conductivity probe. If this works, the steam jacketed conductivity probe should exactly match the actual level inside the drum with no mathematical compensation needed. Problem solved!

Back to top

 

There are several different new technology instruments. Which one works best?

 

Which one works best? That just depends on what you're doing with them. In general, here's the best use for each instrument:

 

Internal DP-Type Instrument (HDSC-DNZ Drum Internal Constant Head Chamber). This is best for hooking up to the DCS for controlling feedwater since it has a continuous 4-20 ma output.

Steam Jacketed Conductivity Probe Instrument is best for hooking up to the automatic boiler protection system. It is a discrete output instrument, and it can be set up completely independent from the DCS so that a DCS failure does not compromise boiler safety.

Steam Jacketed Water Gauge Glass - if you have a regular water gauge glass, do yourself, your maintenance men and your operators a big favor and replace it with this. 

High Pressure Float/ Magnetic Instrument - this is the best instrument we have to offer since it's not susceptible to freezing and less susceptible to leakage. It can install as a reference indicator for the ASME Code boiler to instead one of the DP type instrument or the conductivity probe, but it's not yet available for ASME Code boilers as a control device, pending our work with the ASME Boiler Code Committee. Stay tuned...

Back to top

 

Prove it. How do I know that these instruments really tell the truth?

There are three ways that we know that the new technology instruments are actually telling the truth.

 

Proof #1. The three different technology instruments all substantially agree with one another. This is saying a lot! Traditional instruments are notorious for not agreeing with one another, and the new technology instruments clearly show a much different story than the traditional instruments.

 

Proof #2. It is usually possible, when a unit is down for an extended outage, to enter the boiler drum and look for the water mark which shows where the boiler drum water level controller has been maintaining drum level. Units equipped with the new technology boiler drum level measurement equipment have been documented to have significantly moved their operating water line from the (usually way too high) level before the retrofit to right on the correct normal water level after the retrofit. 

 

Proof #3. This third proof requires a little storytelling. HDSC had one customer which, despite the strong evidence, wanted even more proof. So they funded the design and construction of a research level probe. The probe was a special design conductivity probe instrument that went INSIDE the drum. This obviously took some ingenuity to overcome the technical difficulty of the project. However, if this research instrument could be made to work, it would be easy to know very exactly what the drum level ACTUALLY was inside the drum. After all, a conductivity probe doesn't care about density - it only detects the presence of water or steam. In other words, this kind of a probe cannot lie. HDSC successfully designed and deployed this interesting instrument at the Tongliao Power Plant in China. Sure enough, it corroborated the readings from the other three kinds of instrument very closely. The installation and the results were witnessed by the CSEE (China Society of Electrical Engineers) which is associated with EPRI (Electric Power Research Institute). Of course, the report is in Mandarin Chinese, but this is important enough that we have provided a translation of this document into English, and we would be very pleased if you would take the time to look at it yourself. LINK TO CSEE REPORT 

Back to top

 

What does ASME Boiler Code require me to do?

The ASME Boiler Code has General Requirements for level instrumentation used on power boilers, which are defined in Section 1 of the Code:

o Minimum number of level instruments required for direct reading gauge glasses

o Remote (indirect) level transmitters for observing the level from the control area

o Placement of the indication range for an instrument when installed on the drum

o Minimum pipe size requirements for vessel and drain connections

o Valve requirements, characteristics, and installation requirements

o Information about material and design limitations

The minimum Code requirement for water gauge glasses is to use at least one gauge glass that is always in service for boilers below 400 psi. For boilers above 400 psi, one of the following combinations must be used:

o Two gauge glasses in continuous service and visible by the operator

o Two independent remote indication systems continuously displayed to the operator combined with a single gauge glass, which may be isolated but must be maintained in serviceable condition

o Both of these instruments must be continuously displayed to the operator

o A camera and display for a remote viewing of the gauge glass is acceptable, but two separate means of indication are required

Back to top 

 

What does R-Levels recommend for my boiler/ HRSG? Explain!

It is time to do a little editorializing on the ASME requirements. First of all, the ASME Boiler Code is adamant about one thing - they want a Boiler Gauge Water Level Glass, period. So that much is a given, straight from the Boiler Code. In essence the Boiler Code is not giving any advice regarding how to control the boiler or how to protect the boiler. These are two separate functions (Control and Protection) which need to be considered with any large boiler used in a power plant setting. So, let's examine the situation.

 

We need a drum level measurement device for Control. How many? Let's just follow the logic. If we have one instrument and that instrument fails, our boiler will quickly be in trouble. So what about two instruments? That's better, but if one fails, how do we know which one to trust? That's a he said/ she said situation. So logic leads us to conclude that we need three instruments to achieve Control. The best instrument available for this service is the DP-Type Level Instrument or High Pressure Float/ Magnetic Instrument. 

 

Great. That takes care of Control. What about Protection? Can we just use the same instruments for Control and Protection as well? Yes, but... there are some reasons to shy away from that decision. We need to think very clearly about "common failure mechanisms" and "independence". DP-Type Level Instruments have two glaring weaknesses. One is the possibility of common failure by freezing during cold weather events. Another is the lack of independence. All three instruments rely on the DCS system. If the DCS system has a problem, the boiler is not protected by an independent instrument. That's what the ASME Code means by "two separate means of indication are required". And also some unforeseeable common failure will put the boiler in a very danger position that may rain the unit.

 

So, let's think about Protection. We should obviously use a Conductivity Probe for this function. It is not susceptible to freezing, and it can be made completely independent of the DCS. How many? By the same logic we employed earlier, we quickly arrive at the answer - three. 

 

Step back and consider - this might be "overkill". What if we combine the three signals from the Controls instruments into one signal and use that as one of the three Protection signals? It's not all that likely that we will lose all three DP-Type instruments and both Conductivity Probes in a cold weather event or DCS failure event. So, this brings us to R-Levels final recommendation:

 

We recommend a six instrument solution for any high pressure boiler:

One Boiler Water Gauge Glass (spelled out by ASME Boiler Code) - local indication only or use TV camera to view inside Control Room

Three DP-Type Level Instruments (one "middle of three" signal) - to Control System and one of three inputs to Protection System

Two Conductivity Probe Instruments - to the remaining two inputs of Protection System

High Pressure Float/ Magnetic Instrument  can be used instead one of the DP-Type level instruments or the Conductivity Probe Instruments.

Back to top 

 

How do I get in touch with R-Levels?

Easy. Leave us a message or send us an email and we'll contact you.

 

Back to top