The water quality control for high-pressure steam boilers

Ivan Tikhonov

Almost completely desalinated water is required to be used for feeding steam boilers with the pressure of more than 14 MPa. The salt content of feed water in terms of Na should be no more than 50 mcg/kg, and for direct-flow boilers no more than 20 mcg/kg. The pH value of such water should be equal to 9.1+ – 0.1. At this pH value, the presence of free carbon dioxide in the feed water is completely excluded. However, the feed water may contain bound carbon dioxide in the form of carbonate and bicarbonate which will pass into steam in the boiler and cause corrosion of the steam condensate tract.

For high-pressure boilers, it is customary to use a water treatment scheme with H-OH filters. I.e., all dissolved ions are removed from the water. In this case, carbon dioxide is also removed by the OH-filter. However, the carbon dioxide in the feed water can come from condensate.

To bind carbon dioxide in the condensate and prevent carbon dioxide corrosion, ammonia water is dosed into the feed water. Ammonia reacts with water to form ammonium hydrate (NH₄OH). This increases the pH and electrical conductivity of the feed water. In the boiler, the ammonia evaporates with steam. Carbon dioxide with steam is also removed from the water. When steam condenses, the ammonia in the condensate passes into ammonium hydrate and binds carbon dioxide to bicarbonate according to equation (1). If there is more ammonium hydrate than carbon dioxide, then some of the carbon dioxide will be bound into ammonium carbonate.

NH₄OH+H₂CO₃=NH₄HCO₃+Н₂О    (1)

After dosing the feed water with ammonia, the pH and electrical conductivity increase. Moreover, the value of electrical conductivity will be mainly determined by the concentration of ammonium hydrate.

The question arises, how in this case to determine the concentration of bound carbon dioxide (bicarbonate and carbonate) in the feed water, as well as the concentration of sodium?

To do this, a sample of feed water is taken and its subsequent H – filter is performed. The electrical conductivity of the direct (initial) sample and the H-sample is measured. As a result of the measurements, two values of electrical conductivity are obtained. One value of the electrical conductivity of direct feed water and the second value of the electrical conductivity of the same water, but passed through the H-filter and, accordingly, all cations of the H-filter sample have only hydrogen cations.

Let’s take a closer look at what happens to the feed water after the H-filter.

First, we need to introduce the condition that the feed water does not contain sodium salts, but only contains ammonium hydrate, as well as bicarbonate and ammonium carbonate as a result of the binding of carbon dioxide with ammonium hydrate.

As a result, the electrical conductivity of the direct sample will be determined by three salts (NH₄OH, NH₄HCO₃, (NH₄)₂CO₃).

After the H-filter, all cations (ammonium) will be replaced with hydrogen (H). Accordingly, ammonium hydrate will turn into water, and bicarbonate and ammonium carbonate will turn into carbonic acid.

NH₄OH +НR = NH₄R + HOH

NH₄HCO₃ + HR = NH₄R + H₂CO₃

Depending on the concentration of carbonic acid, some of the carbonic acid will pass into the adsorbed carbon dioxide and, accordingly, it will not be determined as the electrical conductivity of water.

If we take into account that the direct sample does not contain strong acid salts (SO₄, Cl), then the concentration of carbon dioxide in the H-sample can be determined by the concentration of adsorbed CO₂ using the dissociation constant of carbon dioxide at the first stage.

Then

       (2)

where, К₁ = 4,5*10^-7

The concentration of CO₂ in the H-sample in moles will be equal to the concentration of ammonium carbonate and bicarbonate in moles in the direct sample.

Concentrations of ammonium carbonate and bicarbonate separately can be determined based on the constant of hydrolysis of carbonates.

  (3)

where, К_h – Hydrolysis constant, К_h = 0,000213

The concentration of ammonium carbonate and bicarbonate can be determined by setting the NH4OH value. Then, knowing all three concentrations, you can determine the electrical conductivity of direct water (EC_d) by the equation:

where,

EC_nh4on – the value of the equivalent electrical conductivity of ammonium hydrate at infinite dilution (73,6+198,3=271,9 µS*cm²/eq)

EC_nh4hco3  – the value of the equivalent electrical conductivity of ammonium bicarbonate at infinite dilution (73,6+44,5=118,1 µS*cm²/eq)

EC_(nh4)2co3  – the value of the equivalent electrical conductivity of ammonium carbonate at infinite dilution (73,6+138,6=212,2 µS*cm²/eq)

С_nh4on, С _nh4hco3, С_(nh4)2co3  – accordingly, the concentration of hydrate, bicarbonate and carbonate ammonium in feed water, mol/l.

Then you need to check the equality between the calculated electrical conductivity of the direct sample (EC_d) and the actual measured one. If the conductivity values differ, we need to set a new NH₄OH value and repeat the calculation until the calculated and measured values match. It is possible to determine the concentration of NH₄OH based on the measured value of the electrical conductivity of the source water and compare the obtained concentration of ammonium hydrate with the preset one.

As a result of the calculation, the values of ammonium hydrate, carbonate and bicarbonate in a direct sample (in feed water) will be obtained. There is no need to measure the pH of the feed water.

Let’s look at an example.

The electrical conductivity of a direct sample of feed water is equal to EC_d = 4.254 µS/cm. The electrical conductivity of the H – sample is equal to EC_H = 0.198 µS/cm.

Next, we determine the concentration of carbon dioxide in the H – sample.

С_н2со3 = EC_н/394=0,198/394=0,5024*10^-6 mol/l

where, 394 – the value of the equivalent electrical conductivity of carbon dioxide at infinite dilution, µS*cm²/eq.

Then we use equation (2) to determine the concentration of carbon dioxide.

In this case the proportion of carbon dioxide dissociation in the H – sample is 0,5024/0,561=0,895.

Then we set the NH₄OH value to 0.00001 mol / l and calculate the pH value from the hydrate concentration.

Then, using equation (3), we determine the ratio of carbonates to bicarbonate.

Then we determine the value of the electrical conductivity of carbonate (EC_(nh4)2co3 (c))  and bicarbonate (EC_nh4hco3 (c)) using the equations.

EC_(nh4)2co3 (c) = EC_(nh4)2co3  * С_со2 *0,0469=212,2*0,561*10^-3 *0,0469=0,005583 µS/cm

EC_nh4hco3 (c) = EC_nh4hco3  * С_со2 *(1-0,0469)=212,2*0,561*10^-3 *(1-0,0469)=0,0631 µS/cm

Next we determine the electrical conductivity of ammonium hydrate in a direct sample by subtracting the sum of the electrical conductivity of ammonium carbonate and bicarbonate from the measured value of the electrical conductivity of the direct sample.

EC_nh4on (c) = EC_d – (EC_(nh4)2co3 (c) + EC_nh4hco3 (c))= 4,254 – (0,005583 + 0,0631) = 4,185 µS/cm

Then we calculate the concentration of ammonium hydrate based on the calculated value of electrical conductivity.

С_nh4oh (c) = EC_nh4on(c) / EC_nh4on =4,185/271,9 = 0,01539 mmol/l = 0,00001539 mol/l

And, respectively, the pH of the direct sample

As you can see, at the beginning of the calculation, we set the value of ammonium hydrate 0.00001 mol/l (pH=9.0). As a result of the calculation, the values of the set and calculated ammonium hydrate did not coincide. It is necessary to set a new value of ammonium hydrate equal to 0.00001539 mol/l and repeat the calculation.

Having recalculated, we get discrepancies only in the eighth decimal place. Accordingly, it can be said with sufficient accuracy that if direct water contains only hydrate, bicarbonate and carbonate ammonium, they are found in the following concentrations:

С_nh4oh = 0,01538 mmol/l or EC_nh4oh = 0,01538*271,9=4,182 µS/cm м

С_nh4hco3 = 0,00052 mmol/l or EC_nh4hco3 = 0,00052*118,1=0,0614 µS/cm

С_nh4hco3 = 0,0000405 mmol/l or EC_nh4hco3 = 0,0000405*212,2=0,00859 µS/cm

Respectively,

EC_d = 4,182+0,0614+0,00859= 4,252 µS/cm,

Which is almost equal to the measured value – 4,254 µS/cm.

Using this method, calculations were made to determine the possible extreme values of the electrical conductivity of the H-sample. In fact, it was determined when free carbon dioxide begins to appear in the feed water at which conductivity values of H–samples, depending on the conductivity of a direct sample. I.e., all hydrate of ammonium added to the feed water is spent on carbon dioxide binding and only ammonium bicarbonate is present in the feed water and free carbon dioxide begins to appear.

Figure 1 shows a graph of the dependence of the electrical conductivity of the direct sample on the electrical conductivity of the H-sample, at which free carbon dioxide begins to appear in the feed water. It should be said that this graph is valid if there are no strong acid anions in the water (SO₄, Cl).

For example, if the electrical conductivity value of the H-sample is equal to 1 µS/cm, then if the electrical conductivity value of the direct sample is equal to 1.7 µS/cm or less, then the direct water may contain a small amount of free carbon dioxide.

As a rule, the value of the electrical conductivity of the H-sample should be no more than 0.2 µS/cm. In this case, even if the electrical conductivity of the direct sample is only 0.1 µS/cm, such water does not contain free carbon dioxide. Accordingly, if the direct sample conductivity is above 0,1 µs/cm with the conductivity of the H – samples of 0.2 µS/cm, the water definitely does not contain free carbonic acid and there is an excess of hydrate of ammonium.

There is another point to consider. The feed water may contain salts of sulfate and chloride. In this case, to detect their presence in the feed water, it is necessary to measure the pH of the feed water (direct sample).

Figure 1

Figure 2

For clarity, let’s consider the most unfavorable option, when the entire electrical conductivity of the H-sample is determined by strong acids (HCl, H₂SO₄). I.e., after the H – filter, only strong acids remain in the sample. This suggests that the direct sample contains sulphates and chlorides of ammonium or of sodium. The sodium content in feed water should not exceed 50 mcg/kg. In terms of NaCl, this corresponds to a value of electrical conductivity equal to 0.274 µS/cm.

In conditions where the H-sample contains strong acids, carbon dioxide will not dissociate in water and carbon dioxide will be in the adsorbed form. Let’s assume that the amount of carbon dioxide in the H-sample is equal to the amount at which the amount of hydrogen ion, that is contained in the sample as a strong acid, would be formed. In this case the amount of strong acid in moles is equal to the amount of carbon dioxide in moles. So, when determining the ammonium hydrate, it is necessary to subtract the sum of the electrical conductivity of ammonium carbonate and bicarbonate and sodium chloride from the electrical conductivity of the direct sample. As a result, the amount of sodium hydrate as a part of total electrical conductivity decreases and the pH value of the direct sample drops relative to the pH value of the sample without taking into account sodium chloride. Based on the decrease in the pH value we can conclude that there are strong acid salts in the feed water.

In order to determine how much the pH value of a direct sample will decrease in the presence of sodium chloride, several calculations were made. Four values of the electrical conductivity of the direct sample were taken as the initial data (EC_d=5,0 µS/cm; EC_d=2,5 µS/cm; EC_d=0,5 µS/cm; EC_d=0,2 µS/cm). For each value of the electrical conductivity of the direct sample, the pH was calculated taking into account the absence of sodium chloride and the presence of sodium chloride in the sample. In this case, the calculation was performed for different values of the electrical conductivity of the H-sample. Then the difference between the calculated pH of a direct sample without sodium chloride (рH) and with sodium chloride (pH_cl) was calculated. The pH value without sodium chloride is not always greater than with sodium chloride. The calculation results are shown in figure 2.

As you can see, the difference between the two pH values is more than 0.01 units of pH only for small values of the electrical conductivity of the direct sample (0.2-0.5 µS/cm). For large values of electrical conductivity of a direct sample (2.5-5.0 µS/cm) the difference between the pH is less than 0.01 units of pH, which corresponds to the error of the device (pH – meter). That is, the presence of a large amount of hydrate in the direct sample kind of masks the presence of sodium chloride. And in this case, we consider the most unfavorable case, when the entire electrical conductivity of the H-sample is determined by a strong acid. Moreover, when the electrical conductivity of a direct sample is 5 µS/cm, it is observed that the pH of a direct sample with a strong acid anion should be less than the pH without a strong acid anion. This is observed from the value of the electrical conductivity of the H-sample which is more than 1.0 µS/cm. This is due to the conditions that were originally set. With a high value of the electrical conductivity of the H-sample, which is represented only by carbon dioxide, the dissociation of carbon dioxide is about 10%. This results in a much smaller amount of ammonium hydrate in the direct sample than in the presence of strong acid salts.

Keeping in mind the above mentioned, the most reliable and accurate parameter for quality control of feed water is the value of the electrical conductivity of the H-sample. As you can see from figure 2, when the electrical conductivity of the H – sample is less than 0.2 µS/cm, even in the absence of carbon dioxide in the water, which is almost impossible, the concentration of sodium will be no more than 38 mсg/kg.

Thus, the control of feed water by the electrical conductivity of the direct and H – sample is a reliable and fairly accurate method. The value of the electrical conductivity of the H-sample should not be more than 0.2 µS cm, regardless of the value of the electrical conductivity of the direct sample. Control of the pH value of a direct sample is quite effective for determining strong acid anions only if the measurement accuracy of the pH value is more than 0.01 units of pH. The error should be much less than 0.01 units of pH.

 

I hope this information can be useful.