Benchmarking of the operation of Uhle boxes

Abstract

General guidelines for the dimensioning of dewatering at the Uhle box were presented already in the 1970s and 1980s (see reference 2). It was discovered that the dewatering model (DeCrosta) describes dewatering relatively well on average, but differences to actual behavior took place at extreme conditions (very dry and wet felt, high and low dewatering requirements).

A model and benchmarking data for Uhle box dewatering was developed based on felt moisture measurements at several paper mills. Although there are many unknown parameters, a relatively good correlation between the estimation model and measurements could be developed.

1. Background

Dewatering requirements of the PM press section Uhle boxes vary widely depending on the paper grade produced, machine speed, felt position, type and age of the felt, vacuum system design, and many other parameters.

At the design phase of the paper making line, it is, however, important to dimension the Uhle box dewatering capacity properly, because it has a strong effect on the runnability and energy consumption of the line.

Increased energy prices and the development of felts and press sections have also given rise to new challenges in the dimensioning of felt conditioning. In addition, the energy consumption of a PM vacuum system should today be minimized without risking the runnability of the paper making line.

Although the felt moisture content should not always be minimized, a sufficient low moisture level before the nip should be achieved, especially with heavy paper and board grades.

2. Development of the model and benchmarking data

2.1 Felt measurements

Measurement data from about 30 machines producing different paper and board grades were collected and analyzed. Data included scanpro measurements, dewatering measurements, and basic data of the felts (dry basis weight, air permeability, etc.). The results of this data collection are shown at Picture 1.

 

Picture 1: Measured felt moisture before Uhle box and dewatering.

2.2 Analysis of the data, felt moisture

The first observation was that there seems to be a practical limit for the maximal dewatering depending on the felt moisture (Picture 1, red line).

When the felt operated relatively dry (red circle), the relative moisture of the felt after the Uhle box was at a good level of 40% (35 ....45%). Correspondingly, when the felt was running wet (blue circle), the felt's relative moisture after the Uhle box was 50% or more. Also, the mills operating at the area of the red circle were satisfied with the operation.

2.3 Requirements for the Uhle box operation conditions

When the data was further analyzed, it was observed that certain conditions can be set for the operation conditions of the Uhle box to achieve a good level of operation.

A new parameter to describe the operation of the Uhle box was produced, "VacuumTime". This is defined as a product of vacuum level (kPa) and retention time (ms) at the Uhle box (later shortened to VT).

It looks obvious that there has to be either enough vacuum or retention time available to achieve the required dewatering and felt moisture level.

When the dewatering requirements are high (heavy board grades), the VT (VacuumTime) has to be more than 200 Pa*s. It could also be observed that for the machines where the felt was running relatively wet, the VT was smaller (130 ... 200 Pa*s), and the felt's relative moisture after the Uhle box was higher (over 50%).

It should also be observed that when the press section and felts are designed for nip dewatering, the requirements for the Uhle box seems to be very low, especially for paper grades and low basis weight liner and fluting grades (see articles 2 and 3).

Based on this data, a basic rule can be stated:

- VT (Pa*s) should be more or less the same as the required dewatering (g/m²) at the Uhle box.

2.4 Further tests and trials

More tests were conducted to determine the effect of retention time and vacuum on the felt dryness and dewatering at the Uhle box. These tests were carried out at relatively low dewatering level (15 .... 50 g/m²), thus the VT was also lower (see Picture 2).

Picture 2: Felt moisture before Uhle box as a function of VT (VacuumTime).

Felt moisture decreases almost linearly as a function of VT when the felt moisture level is over 700 g/m², but when the felt moisture further decreases (relative moisture under 45%), the curve saturates and the increase of VT no longer has any effect at all.

It can even be concluded that when the dewatering requirements at the Uhle box are small (low basis weight paper grades or high nip dewatering), vacuum time (VT) of 40...80 Pa's should always be enough. The exception is pick-up felt where dewatering can be higher and 100...120 Pa*s is needed at least with new open felt.

When the felt compresses (open volume decreases) and nip dewatering starts, the requirements for the Uhle box can be very low.

3. Modeling of the dewatering at the Uhle box

3.1 KG-dewatering model

Because a certain logic in the measurement data and trials could be observed, a black box model for the estimation of dewatering at the Uhle box was developed. Dewatering was estimated based on the following four parameters:

  1. Retention time (dwell) of the felt over open area of the Uhle box (ms),

  2. Vacuum of the Uhle box (kPa),

  3.Felt moisture before Uhle box (g/m2),

  4.Felt air permeability (cfm).

 

Felt moisture was selected as one parameter because dewatering looked to correlate strongly with it. Also, felt air permeability was taken into account mainly because it has a clear effect on the air flow through the felt (see article 5), and thus it could also correlate with dewatering.

3.2 Correlation of the model and measurements

The estimated dewatering with the KG dewatering formula is plotted against the measured data in Picture 3. The correlation coefficient is 0.67.

The average error is 42 g/m2, and the estimated dewatering is practically always inside +/- 75 g/m2 compared to the measurements.

 

Picture 3: Correlation of the estimate to the measurements.

4. Discussion

As the visual review demonstrates, the variation is still pretty large due to, for example, the felt aging and other parameters, such as felt type. However, it gives a rough estimate for the required VT (Pa*s) as a function of the required dewatering and felt moisture level.

It can also be seen that the dwell time (ms) and vacuum level (kPa) compensate for each other at a certain level. However, some limitations exist:

- The vacuum level cannot be decreased too much, and it should be over 25 kPa.

- If the dewatering requirements are high, it is difficult to get the relative moisture after the Uhle box much lower than 40%, which could be used as a good target moisture level.

No clear difference between slot sizes or cover geometry could be seen in this study.

5. Examples

Even though the felt moisture or dewatering cannot be estimated exactly, the required typical vacuum level and dwell time can be roughly estimated with the KG dewatering formula.

For example, if the dewatering level of 180 g/m2 is targeted at the felt moisture level of 900 g/m2 (before the Uhle box), the VT should be about 200 Pa*s (see Picture 4).

The required dewatering can be achieved with several combinations of vacuum level and dwell time (vacuum from 30 kPa to 50 kPa, and retention time from 4 ms up to 7 ms).

Picture 4: Estimation of dewatering.

For low basis weight products (or machines running mostly based on nip dewatering), the requirements for the Uhle box are different. They are mainly used for cleaning the felt and equalizing the moisture profile (e.g. removing the moisture peaks of showers); the practical minimum level of VT in these cases is about 40 Pa*s.

6. Application possibilities

The KG dewatering model provides one opportunity to do benchmarking (check the operation conditions) of the operation of Uhle boxes. For example, in the event where felts are running wet, it can be used to check if the vacuum level and dwell time are at the required level.

Also, in energy improvement projects, saving potential can preliminarily be estimated based on the model; for example, if the felt conditioning is over-dimensioned (see articles 1, 4 and 5), or the felts are runng too dry.

7. Summary

Even though felt moisture and dewatering can be roughly estimated, felt aging and other parameters still have a significant effect on the operation conditions.

For this reason, the vacuum system should have a wide operation window, and the vacuum level should be controllable without losing the energy efficiency of the system.

In the event that vacuum system efficiency and controllability are not good, a vacuum system study (pre-engineering) can always be recommended to determine the reasons for the efficiency losses and improve the controllability of the system.

 

Related articles and references:

1. O. Kaapa, 18-19 / 2009 Wochenblatt für papierfabrikation: 

    Drastische Einsparungen in der Pressenpartie mit Auswirkungen in die Trockenpartie.

2. Edward F. DeCrosta, TAPPI May 1980, Vol 63 No. 5: 

    Air flow requirements for conditioning press felts at suction pipes

 

3. O. Kaapa and co , 2/2013, Wochenblatt für Papierfabrikation, 

    Vakuum – Kapazität in der Pressenpartie.

4. K. Kokkonen, Results pulp&paper No 1/2011: 

    Energy savings through a new vacuum system concept.

 

5. K. Kokkonen, Results pulp&paper No 3/2012: 

    Curbing energy costs through a blower system rebuild.

 

6. K. Kokkonen, Wochenblatt fûr Papierfabrikazion 2/2012, 

   Energieeinspar-potentiale bei der Filzkonditionierung,

Effect of seal water temperature to vacuum pump capacity

Abstract

Luquid ring pumps ( later LRP) are widely used at paper mill vacuum systems. They are known to be very reliable, but sometimes capacity losses ( efficiency losses) can take place because of wearing of the pump, traditional vacuum level control, lack of seal water or too high temperature of the seal water.

Air flow measurements were carried out at a paper mill wet end vacuum system, and results were compared to pump characteristics and  theoretical estimation of the capacity loss due to the high seal water temperature ( seal water cooling tower at the mill had been bypassed).

Measurements corralate well with the theoretical calculation, and electric energy saving potential could be identified ( if the seal water temperature is first controlled down to the normal level).


1. Open seal water system

So called open seal water system is often used at the PM vacuum systems. This open system is very common at Scandinavia and northern parts of America, where the fresh water consumption is not limited.

Often the seal water temperature is controlled with cold fresh water make-up, and typical setpoint is about 30 C ( about 90 F). If water separation is in a good condition, the seal water is not contaminated and can be directed back to environment without further threatment.

 

                                         Picture 1: Open seal water system

2. Closed seal water system

In many areas cold fresh water resourses are limited, and the seal water system needs to be closed. About 10 % make-up water can be recommended also at closed systems.

Good desing temperature of seal water is about 30 C, but at warm climate it is not always possible, and higher operation temperatures has to be accepted ( up to 35 C ... 40 C ). 

Water separation is needed also at closed systems to avoid contamination of the seal water loop and plugging problems of the cooling tower. This is often the reason, why the cooling towers are not operating porperly, and have even been bypassed.

                                   Picture 2: Closed seal water system


3. Effect of the seal water temperature to vacuum pump capacity

Seal water temperature has a clear effect to the capacity of LRP ( compare for example to the link: http://www.dekkervacuum.com/static4/service.asp  and picture 3 ). 

 

                           Picture 3: Vacuum pump capacity and operation temperature


4. Process temperature

The temperature of the incoming air to to vacuum pump is typically about 37 C ( from 30 C up to 40 C). Higher air temperatures ( over 40 C) can be measured only at the beginning of forming section, where air flows are small and air has contact to the warm white water.

Sometimes the PM headbox temperature is over 50 C, but the air flow through the web and felts however cool the process, and also the paper web temperature at press section is normally under 40 C.

Thus for exampe 37 C ( 100 F) is typically a good design temperature for the incoming air to the vacuum pumps. A model was developed to estimate the effect of seal water temperature to the pump capacity.  Calculation is based on change the water vapour pressure at different temperatures, and it seems to correlate well with picture 3.

                                   Picture 4: Water vapour pressure
 

5. Air flow measurements at the paper mill

Benchmarking of the PM of the vacuum system indicated high SEC ( Specific electric Energy Consumption), in average about 100 kWh/tn. Compare also to the blog, Oct 2012. Vacuum system study was carried out at the mill, and air flows from the PM and temperatures were measured.

This mill has a closed seal water system ( compare to the picture 2), and for some reason the cooling tower was bypassed. Because of the high process temperature and warm climate, incoming air temperature to the vacuum pumps was at high (about 40 C) and pump surface temperatures about 50 C. The seal water temperatures were over 55 C ( up to about 60 C).

Measured air flows were compared to the vacuum pump design figures, and are presented at the picture 5. Red and yellow curves are calculated with the model, and red dots are the measured air flows from different positions of the paper machine. 

                                   Picture 5. Pump capacity compared the desing.


6. Discussion

Air flow measurements were done with pitot pipe, thus accuracy in mill conditionds can not be very good. However, the measurements and the capacity estimation cleary seem to correlate with each other (picture 5).

High seal water temperature decreased the pump capacity at high vacuum positions (vacuum over 50 kPa) in average about 20 %.


7. Energy saving potential

In this example mill the NRL load of all vacuum pumps is about 4,0 MW ( installed motor load about 5 MW).

If the seal water temperature is decreased by 10 ... 15 C, capacity of the pumps increase and the pump speeds could be correspondingly decreased to save energy. In practice rotation speed optimization is needed at high vacuum positions only. Alternatively vacuum levels at the PM will increase ( = if pump speeds or vacuum level controls are not updated).

Already 10 % capacity improvement in average equals to 400 kW savings which equals annually about 200.000 € savings ( energy price 60 €/MWh). 

However, the first action would be get the seal water system and water temperatures under control. Only after it, further optimization can be considered.


8. Summary and conclusions

Water vapour pressure increases exponentially when the temperature increases, thus the vacuum pump seal water temperature should always be under 40 C (favourably at 30 C). 

High seal water temperature ( over 40 C) is harmfull especially at high vacuum pumps ( = vacuum level over 50 kPa).

Water separation must always be adequate, to avoid contamination of the seal water.  
 

References and links
  
     1. Effect of seal water temperature to the vacuum pump capacity
          http://www.dekkervacuum.com/static4/service.asp

     2. Estimation of pump capacity on-line as a function of seal water temperature
         www.kgu.fi ( these pages are under consruction)
        

 

Energy efficiency of PM vacuum system equipment

 

1. General

There are many reasons, why the electric energy consumption of the PM vacuum system 
varies from one machine to another:

     1. Paper machine geometry

      The amount of suction rolls and suction boxes vary from one machine to another, 
      and the required vacuum level differ depending on the machine geometry.

     2. Vacuum system efficiency
     
     The design of the vacuum system process and vacuum level control  may 
     cause efficiency losses at the system ( bleed air or expansion losses). System
     efficiency losses take place, when the capacity of vacuum pumps or blowers do 
     not match  with the vacuum level or air flow from the machine.

     3. Efficiency of  vacuum system equipment

     The efficiency of equipment itself varies also. At the following is presented one
     example to evaluate the efficiency of vacuum system equipment ( = fans,  pumps 
     and blowers).

2. Specific energy consumption

The required electric energy consumption of fans, pumps and blowers can be presented
as a function of the vacuum level ( compare to the picture 1).

The attached graph has been developed based on the best available technology at 
different vacuum levels, and can not be achieved with one single piece of equipment.
at the whole operation area ( low and high vacuum lelvels ).



   
       Picture 1: Specific energy consumption


The picture 1 presents a simple method for estimating the efficiency of equipment itself, 
and it does not take into account the system efficiency losses.
     - Specific energy consumption is estimated based on the suction conditions of the
        blower or pump ( air flow and vacuum level at the intake ).  

Based on picture 1, a "rule of thump" has also been developed, to estimate the energy
consumption.

      P ( kW) = Q ( m3/s) * p ( kPa) * K                                                    ( 1 )

      Where K = constant, 1,3 ... 1.5

3. Elecric energy consumption and efficiency

Opinions of different vacuum systems are often presented, where one or another system 
is found to be better compared to the other ones. However, if the comparison is done 
based on the electric energy consumption, the result depends on many variables.
The process desgn and control strategy plays also an important role in it.

Overall feasibility of vacuum system improvement ( or new line investment)  depends also
on investment cost, required space ( mill lay-out) and heat recovery. possibilities. Some general quidelines can however also be presented based on electric energy consumption.

   1. At low vacuum positions ( < 30 kPa ) of the PM, a simple fan should always be
       considered first. The efficiency of liquid ring pump is always poor at a low vacuum, 
       and the invesment cost of a simple fan is smallest.

   2. At medium vacuum area ( 40 kPa ... 55 kPa) the the process system design plays
       an important role.  Conrollability must be good when the requirements of the PM vary
       depending on the produced grade, PM speed and felt age. 

       The efficiency of pumps is also acceptable, if the rotation speed is not too high.
       
       Multistage and singlestage turboblowers normally operate at at good efficiency at 
       this vacuum level. Important is to be able to control vacuum and  air flow according 
       to PM needs ( diffusor and variable speed drive needed to avoid system efficiency 
       losses = bleed air and expansion).

   3. High vacuum ( > 60 kPa) can be achieved only with multistage blowers or liquid ring
       pumps. At well designed systems the diference at SEC is reasoably small.

      Sometimes pumps at the older mills have been selected to operate at high speed, 
      and the efficiency can be very poor.

      Multistage blower normally operates at a good efficiency level. However, because of 
      over dimensioning , or variating operation conditions, expansion may  take place 
      ( = vacuum too high at the suction ) and reduce the overall efficiency of the system.
      On the other hand, the multistage blower offers the possibility  for the heat recovery
      ( exhaust air at high temperature).

4. Summary

In general (especially at energy improvement projects) it can not be decided without 
a feasibility study, which system is the best one. 

Even though fans and blowers often have better  efficiency ( low vacuum area), quite 
often 80 % of the energy savings can be achieved with 20 % investment ( 20 / 80 rule),
when the operation of the existing system is optimised and controls are updated 
( instead of replacing them with new pieces equipment).

At existing systems the first action is always to make the benchmarking to find out 
weather the system energy consumption high or low compared to the other machines
( compare to the earlier article, block fall 2012, Benchmarkng ).

This article offers an alternative to find out the reasons for the high energy consumption
at the vacuum system. Comparison of equipment efficiency often requires a more detail
pre-engineering ( air flow and vacuum level measurements at the machine)


                                                                                                         Kari U Kokkonen 
-------------------------------------------------------------------------------------------------------------------------