Thermal fluid versus steam: a CAPEX-OPEX comparison

Steam versus thermal fluid

It is becoming increasingly common to see the expressions CAPEX and OPEX used in commercial transactions. Both expressions derive from English-language contractions, CAPEX for capital expenditure and OPEX for operational expenditure, and can be integrated, with small variations, into expressions for implementation costs and operating costs that in years gone by, were decisive in decision-making prior to acquiring industrial equipment.

Indirect heating installations are those in which an intermediate fluid is used between the heater or boiler and the consumer appliances. This fluid, which circulates in a controlled manner, receives the energy generated in the boiler and transports it to the consumer appliances, to which it transfers it.

The two most common fluids used in indirect heating installations are water, generally steam-based, and so-called thermal fluids or oils. The choice of intermediate fluid used determines the type of boiler, since each type of fluid requires a specific type of heating and installation equipment.

Intuitively, water in the form of steam may seem to be the most suitable intermediate fluid as it is easily obtained at very affordable prices. However, as we will see below, the characteristics of this fluid mean that we have to consider water as a “false” friend in its role as an intermediate fluid for transporting energy in industrial installations, as it has many deficiencies compared to thermal fluids, deficiencies which have a direct impact on CAPEX and OPEX.

It is also commonly accepted that the CAPEX of a steam installation is lower than that of a thermal fluid installation. That is indeed the case, but this statement must be qualified. It will occur when the boiler and the steam installation are at the so-called “basic level”, which involves two-step boiler smoke – although to add confusion, the term “three steps of smoke, two in chamber” is sometimes used, referring to a condensate tank without a steam connection that facilitates deaeration, without taking advantage of boiler purges, and so on.

Obviously this type of installation will work and will have a very low initial CAPEX, at a clear cost of a detrimental energy efficiency (that of a complete steam installation is already much lower than that of a thermal fluid one, as we will see later), a shortening of the useful life of the boiler (two smoke steps or “three, two in chamber” imply overheating by a specific thermal load in very high-combustion chambers and therefore frequent repairs) and poor OPEX and CAPEX in the following years on account of to having to pass on high rates of amortisation.

It must also be taken into account that the total cost of the thermal fluid load is usually included in the CAPEX of a thermal fluid installation. It must be borne in mind that this cost should be passed on within at least 10 years, which is the minimum life of a thermal fluid load in a proper installation, or the costs of water and its treatment should be included throughout 10 years in the CAPEX of the steam installation.

Therefore, this favouritism of steam installations in implementation costs must be analysed in detail. In any case (and considering it valid, although not as obviously so as it may sometimes seem), the following table readily shows us the strong points – the positives – of each type of installation with respect to the costs of implantation or operation. In the following pages, we will explain these positive points and examine their causes.

Thermal fluid Steam
(Cost of installation)
Service life Longer Shorter
Updates Flexible Less flexible
Purchase cost More expensive Cheaper
(Cost of operation)
Maintenance Lower (fewer purges and no corrosion) Higher (frequent purges and corrosion)
Boiler room No Mandatory
Manpower needed No operator required Operator is required
Spare parts stock Smaller Bigger
Fuel consumption Lower Higher
Electrical consumption Higher Lower
Efficiency Higher Lower



The phenomenon of encrustation – the formation of hard deposits on metal surfaces – manifests as disruptions in the carbon balance and precipitation of magnesium and calcium carbonates that are diluted in the boiler feed water.

The encrustations act as pseudo-insulators, since the thermal conductivity of iron is 40 cal/h mºC. whereas in calcium carbonate (CO3Ca) it is 6.3 cal/h mºC. The encrustations act as pseudo-insulators, since the thermal conductivity of iron is 40 cal/h mºC. whereas in calcium carbonate (CO3Ca) it is 6.3 cal/h mºC. The possibilities of transmitting the energy to the water decrease and part of it goes to the combustion gases that are evacuated through the chimney. There is therefore a decrease in boiler efficiency and an increase in fuel consumption.

The importance of the characteristics of both the water used to feed the steam boilers and the water inside the boiler is such that in the EU they are regulated by the obligatory EN-12953-10 standard. We must keep in mind that due to the porosity of encrustation, water in contact with superheated steel can cause corrosion – as we will see later – when nascent oxygen appears. If a substantial piece of encrusted material is detached, violent vaporisation of the water can occur, with the danger that this entails.

Therefore, corrosion must be minimised for reasons of both energy efficiency and equipment safety.

Scale problems require water treatment with decalcifiers, additives, etc. to ensure acceptable feed water and increased preventive maintenance. These operations are essential to moderating the harmful consequences of encrustations, which can never be completely eliminated.

The treatment of feed water has an impact on both CAPEX – investment in equipment – and COEX – additives, salts.

In addition, although the boiler is fed with treated water, it contains a large quantity of mineral salts, since the decalcifier merely carries out an ionic transformation which prevents the salts dissolved in the water – calcium carbonate and sodium carbonate – from adhering to the walls of the boiler, partially avoiding encrustations. As they do not adhere to the walls of the boiler, these mineral salts accumulate at the bottom of the boiler, forming so-called “sludge” which must be removed by boiler purging.

These salts reduce the steel’s thermal transmission to the water, which in turn leads to an increase in the temperature of the home and of the boiler tubes, which can lead to a loss of mechanical strength and a considerable reduction in the performance of the equipment.

Periodic purging of the boiler so that it has an acceptable useful life implies a significant loss of energy in these purges and therefore additional fuel costs – COEX -.

To get an idea of the magnitude of energy losses due to boiler purges, consider a boiler of 2000 kgv/h at 6 bar pressure, with feed water at 250 ppm – 0.25 gr/litre – of total alkalinity.

The accumulation of impurities in one hour: 0.25 gr/litre x 2000 kgv/h = 500 gr/h.
In an 8-hour workday, there are 4 kg of impurities: 8 x 0.5 kg/h

To determine the quantity that should be purged, there are several expressions from steam boiler and accessory manufacturers, all valid and very similar.

The expression we will use is found in Steam in Industry, by Spirax Sarco and in Water in Steam Boilers, by Ygnis.

A recommended salt value inside the boiler is between 2000 and 4000 ppm, depending on the size of the boiler. For this calculation we will use the mean value, 3000 ppm.

The amount of boiler purge to perform will be:

i n p u t   p p m   x   p r o d u c t i o n d e s i r e d   p p m     i n p u t   p p m = 2 5 0   ×   2 0 0 0 3 0 0 0     2 5 0 =   1 8 1 k g / h

We see then that the boiler purge amount to be carried out is on the order of 9.1% of the nominal production of the same.

Given that the purges are not steam, but water in the liquid phase at the service temperature of 158, 83 ºC according to Table 1 on page 5, the percentage of energy loss is less than 9.1% of production, given that the energy for such production includes the heat of vaporisation.

The attached Graph 1 shows the energy efficiency according to the percentage of boiler purges. The 9.1% purges over the total production of the equipment represent approximately 3-3.5% of lower energy efficiency.

Energy efficiency and purges

Chart from Water in Steam Boilers, by Ygnis

In order to avoid this significant loss of efficiency, it would be logical to carry out better treatment of the feed water and to reduce operating costs. However, this would require very sophisticated equipment – meaning increased implementation costs – and exhaustive monitoring to conform their proper functioning, so that the reduction in operating costs would not be that significant.

Typically, boiler manufacturers do not consider more sophisticated feed water treatments until purge values reach 25% of nominal production. This value, like all those in this chapter, are those recommended by Spirax Sarco, Ygnis, etc.

In our hypothetical case, the energy efficiency would be approximately 84.5%. These losses of 15.5% exclusively include the efficiency losses from boiler combustion and boiler purges. We must therefore add losses which are due to the “change of phase”

In all calculations, 100% condensate utilisation has been considered. If the percentage is lower, the quantity of boiler purges to be carried out would obviously be lower, as the quantity of treated – and therefore more adequate – feed water would be greater. However, the loss of energy would be much greater, because as we have seen before, the energy losses due to “phase change” would increase exponentially.

Although the useful life of steam boilers will be dealt with in greater depth on page 9 of this report, we stress here that even with all these precautions – boiler purges, water treatment and so on – the useful life of a steam boiler is significantly lower than that of a thermal fluid boiler, since these operations merely minimise the presence of encrustations inside the boiler and can never eliminate them absolutely, and a further determining factor at this point is the corrosion suffered inside the steam boilers.

Phase changes

A steam installation uses liquid-phase water in the boiler feed process. It converts it into a vapour phase inside the boiler and is thus maintained during the process of transporting the energy and until its transfer to the consumer appliances. In consumer appliances, when the transported energy is transferred the water in the vapour phase condenses into a liquid phase and, through the condensate network, returns in that phase to the boiler.

Inside the boiler, we provide the so-called “vaporisation heat” or “vaporisation enthalpy” which will later be delivered to the consumer appliance, PLUS the precise heat to transform the feed water into water at the working temperature. In the case of a boiler at a working pressure of 6 bar, this temperature is 158.83 ºC – see Table 1 attached -.

After the condensation experienced by the water vapour in the consumer appliance, we obtain water in the liquid phase at the temperature previously indicated. If it were possible to reintroduce all the condensate produced in the consumer appliances and drains that a steam installation also requires at the same steam distribution temperature – service temperature -, the losses due to a phase change would be non-existent and the system could be compared perfectly with a thermal fluid installation, where as there is no phase change, the losses due to this purpose are null.

Table 1. Properties of saturated steam

However, having a condensate network pressurised to the operating pressure – which would make it possible to maintain the temperature – is impossible. The steam traps and drains of steam installations – at the end of the branch, in each consumer appliance, at low points, in straight sections -, although they operate with differences in temperature, liquid phase/gas phase densities or velocities, their dimensioning is directly proportional to the differential pressure. Minimising this differential pressure would involve large-size traps and a fully pressurised condensate return network, including a reservoir. Valves and pumps should also be able to withstand operating pressure. The CAPEX would be very high, impossible to assume and the OPEX would also be considerable, as the sophistication of the system would require strict maintenance.

In large steam systems, a moderately pressurised condensate network is chosen. With operating pressures on the order of 20-30 bar – between 210 and 230 ºC -, the condensate network is approximately 6-8 bar – between 158 and 170 ºC -. There is a significant energy loss – the energy that must be supplied by the fuel to raise the water in liquid phase from 158 ºC to 210 ºC is not recovered – but at least partial use is achieved.

Facilities with lower working pressures opt not to pressurise the condensate network, with the objective being high condensate recovery. Not pressurising the condensate network implies that the maximum temperature of the condensate upon arrival at the condenser reservoir is 100ºC, being in practice between 80-90ºC. Obviously energy losses are significant, as we shall now see.

The extreme case is when, as in many steam installations, condensate is not used or the percentage of condensate recycling is less than 50%. In terms of energy, they are devastatingly wasteful installations.

This leads to significant increases in fuel costs, as energy must be supplied which is NOT SUBSEQUENTLY CEDED to the appliances.

The graph shows the cycle of 1 kg of steam at 6 bar pressure.
Energy cycle of 1 kg steam at 6 bar pressure

Graph 1. Energy cycle of 1 kg steam at 6 bar pressure

The energy efficiency of the entire indirect steam heating system from a kg of steam to a bar pressure is one:

ρ = 2 0 8 6 k J / k g   ×   1 0 0 2 8 6 4 k J / k g = 7 2 . 8 3 %


Combustion. Yield 88% 12%
Losses due to boiler purges 3.43%
Phase change losses 11.73%

When analysing the same cycle for a thermal fluid installation, we see that the differential points are the boiler purges and phase change losses. In both cases, losses by radiation from the pipes have been neglected, since with correct insulation, they must be less than 0.2% in both cases.

In order to deliver the same energy as in the steam cycle – 2086 kJ/kg -, the only losses suffered by the thermal fluid installation are those resulting from boiler combustion, to which we may assign the same performance as in the previous cycle – 88% -.

Thermal fluid installation energy cycle

Graph 2. Thermal fluid installation energy cycle

The percentage difference in energy efficiency between the two installations is:

88% – 72.83% = 15.17%.

In order to achieve 2000 kgv/h, the steam system will need 2000 kgv/h x 2864 kJ/kgv = 5728000 kJ/h. From this energy, consumer devices will receive 4172000 kJ/h.

To satisfy the same consumption – 4172000 kJ/h -, the indirect thermal fluid heating system only has to provide 4172000/.88 = 4740.909 kJ/h. This means a difference of 987090 kJ/h.

This difference corresponds to 24.38 Nm3 of gas or 23.12 kg of diesel, for EVERY HOUR that the equipment is in operation

With a natural gas ICP of 40474 kJ/Nm3(1) and 42700 kJ/kg (1) of diesel, and with approximate prices of €0.036kWh for natural gas – equivalent to €0.4212 /m3 – and 0.89€/l – equivalent to €0.98/kg – for diesel, we would be talking about differences of between €10.26 and €20.57, EVERY HOUR.

These values are approximate as fuel prices may vary slightly depending on the contracted tariff, supply company or time of year, but they do signify an order of magnitude of the economic differences in fuel that occur depending on the type of installation.

(1) Source Eurosta, IEA and Resolution of the Secretary of State for Energy of 27 December 2013 modifying Order ITC/2877/2008

Electricity consumption

Within this section on energy efficiency, we will also consider the electricity consumption required by both types of installations.

Based on the fact that the electrical consumption of the burner and of regulators and automatic features is practically the same in both installations, we will focus on evaluating the consumption of pump motors on a basic level. This consumption of burner and automatic features can be evaluated to be 4 kWh.

In a thermal fluid installation, we must consider the consumption of the general recirculation pump, which, unlike the pumps of a steam installation, works continuously while the installation is in service condition. There are other pumps in the thermal fluid installation, such as the filling-emptying pump or the secondary circuit recirculation pumps. We will not consider the consumptions of the same, since in the first case, it is only put into operation for tasks involving filling and emptying the installation, an event that happens only very occasionally – normally, years can pass without this pump being used-.

With regard to secondary circuit recirculation pumps, we will not consider their consumption as they are part of the (optional) achievement of quality requirements in the finished product, which a steam installation cannot achieve. It is our intention to compare installations with similar service conditions.

A thermal fluid system of similar performance to a steam system of 2000 kgv/h at 6 bar is equipped with a recirculation pump with a 20 hp – 15 kW – motor. The consumption will be around 15 or 16 hp 11.5 – 12 kW -.

The hourly consumption in the thermal fluid installation would be 12 kW + 4 kW = 16 kW.

The price of the industrial kW can vary according to supplier and tariff, but we can evaluate it approximately between 0.088 and 0.12 €/kWh. Considering the higher price, the cost in electricity of the thermal fluid installation would be 16 x 0.12 = €1.92/h.

In a steam system there are basically two pumps: the boiler feed pump and the condensate return pump. Its operation is related: the more or less condensate return there is (and therefore the more frequently the condensate return pump operates), the less or more the boiler feed pump will operate.

In some installations, the condensate return pump employs a mechanical float action and does not need an electrical power supply. In any case, as we are going to see now, the consumption of both pumps is very low.

The feed pump of a steam boiler of 2000 kgv/h at 6 bar will have a motor of approximately 1.5 kW, with condensates (assuming it is not mechanical) of 1 kW. We can consider that the sum of the two working will be approximately 75% of the time. Therefore, the expected consumption is 0.75 x 1.25 = 0.94 kWh, with a cost of 0.92 x 0.12 = €0.1104/h.

The steam installation has a favourable difference in electrical energy operating costs of €1.8096/h.

Although it is a fact to consider, we should remember that as we have seen on the previous page, the thermal fluid installation has a favourable difference in fuel operating costs between €10.26 and €20.57.

From an energy efficiency point of view, there is no doubt that the thermal fluid installation has a VERY FAVORABLE OPEX.


It is evident that the useful life of the equipment, whether boilers or installations, is a point of vital importance in the evaluation of both CAPEX, where we must consider the rates of amortisation of the equipment, and of CAPEX, where the costs for repairs, maintenance and even management of stocks of spare parts have their importance.


Corrosion is a process whereby metal, in contact with its environment, tends to change from a form of pure metal to a more stable form. The steel is gradually dissolved by the water and oxidised by the oxygen it is carrying, forming oxidation products based on iron oxides. This process occurs more quickly in steam boilers, since at high temperatures, with the presence of corrosive gases and solids dissolved in water, corrosion processes are stimulated.

In a steam boiler, even with properly treated water, different types of corrosion occur that affect different parts of the equipment. We will thus observe a so-called general corrosion that tends to dissolve or attack the metal in a uniform manner, being overheated by steam. It is essentially produced on the surface of the smoke tubes. A particular case of this type of corrosion is due to the acidity of the water, as iron and other metals dissolve more quickly the lower the pH value, the more acidic the water. Hence the need to maintain a strongly alkaline pH – a high one – inside the steam boilers and during the boiler purges we have seen before.

Corrosion by oxygen or pitting. When a metal is irregularly covered by mud and encrustation, in the presence of dissolved oxygen. Fully covered areas are most commonly attacked and become the centre of localised, deep corrosion, visually resembling black tubers.

Although in a global computation there may be significant losses of material, the gravity of this type of corrosion is that being very localised attacks produces perforations and pitting in the equipment and more or less certain areas – tubes, home – that necessarily require stops and repairs to be critical points in the operation of the equipment. In extreme cases, significant corrosion can lead to replacement of the boiler.

Caustic corrosion is produced by a local over-concentration in areas with a high thermal load – such as the inversion chamber or the firebox- of alkaline salts such as caustic soda. This type of corrosion manifests itself in the form of deep cavities, similar to pitting.

Carbon dioxide corrosion occurs when part of the CO2 is partially diluted in water in the form of a gas, while another part reacts with the same CO2 to form carbonic acid, which is partially dissociated into carbonate or bicarbonate ions. These ions will attack the material. This corrosion is mainly evident in the condensate return lines and has effects on the boiler, as the acids produced are dragged into the boiler along with the feed water.

Obviously, in a thermal fluid installation, there are no corrosion problems. Let’s not forget that a lubricating oil does not corrode, and has no carbonates or salts that can precipitate. There are therefore no decalcifiers or additives needed, no losses due to purging, and most importantly, thermal fluid boilers and installations enjoy long service lives with no need for repairs and maintenance against corrosion.


Pressure/temperature ratio

The temperature at which water in liquid form becomes vapour – the gas phase – is called the boiling temperature or the vapour saturation point and is directly related to pressure. Thus, at atmospheric pressure, steam is obtained at a temperature of 100 ºC. If the needs of our manufacturing system should require higher temperatures, it is necessary to have a higher pressure.

Pressure/Temperature Graph. Thermal fluid vs. steam

Fig 1. Pressure/Temperature Graph. Thermal fluid vs. steam

Fig.1 shows the Pressure/Temperature graph of both water and a standard thermal fluid. If our manufacturing system requires a working temperature of 159 ºC, with water as an energy transport fluid, we must work at a pressure of 6 bar.

With the thermal fluid, there is practically no such relationship and the pressure requirements are limited to the pressure losses typical of the installation due to friction in pipes.


The independence of pressure from temperature in the thermal fluid also makes it possible to completely eliminate explosion hazards and, according to the current Pressure Equipment Regulation, allows installation of thermal fluid boilers in any work room, without the need for a boiler room, with savings on the civil works that a new installation can represent, or the space gains that it can produce in an industry that is already in operation.

Another advantage that this section may represent is the possibility of locating the thermal fluid boiler very close to that of the consumer appliances, enabling short and therefore economic installations.

Another point that corroborates the greater safety and simplicity of thermal fluid installations is the fact that according to current regulations, it is not necessary for the personnel in charge of it to have boiler operator licenses, unlike with steam installations where there must be at least one licensed operator per shift.

Obviously it is not a big difference in operating costs, but it should be taken into consideration.


A thermal fluid installation offers simplicity as standard. A boiler, a pump, a distribution network and an expansion tank. Its general configuration is not very different from a home heating network. There are no traps, condensate network, feed tank, condensate tank, salt tank, deaerators, steam network for oxygen removal or boiler purges.

Not only is daily activity operationally simple, with – as we have seen before – no need for execution by specific personnel, but this simplicity makes the costs of preventive maintenance virtually non-existent.


In a steam installation, limitation of the working pressure is absolute. If you have a 6 bar pressure boiler and due to some new circumstance, for example a new product line, you must work at a temperature of 200 º C, a change of boiler to a new pressure of 16 bar is required in all cases, as well as a redesign of the pipeline network in almost all cases. The valves must also be changed most of the time. The opposite case, working at a lower temperature/pressure, requires a new pressure test and/or replacement of safety valves, and in most cases re-dimensioning of the general network.

In thermal fluid installations, flexibility is absolute, with no need for changes in either the boiler or the installation. It is possible that such an advantage will never be exploited, but if there were a need for it, the difference in the implementation costs of the new project would be abysmal.

From the point of view of safety and flexibility, the advantage of a thermal fluid installation is evident in both CAPEX and OPEX. However, they may not be significant values. If there is no provision for modification of the production process, and no significant changes therefore need be made to the installation, the absolute flexibility of the thermal fluid installation does not entail an advantage and does not affect the CAPEX

Likewise, if due to machinery distribution needs the location of both boilers will be the same or if there is already a boiler room, the less demanding regulatory requirements for thermal fluid boilers will result in relatively low savings on civil works. On the other hand, if it is necessary to build a new boiler room for the steam boiler, the possibility of installing the thermal fluid boiler in a location close to the consumption points, and without the need for specific walls, can mean an important CAPEX saving.

The OPEX savings that may not require a boiler operator with a specific license will obviously be low. Curiously, in general, the greatest advantage for thermal fluid installations of not needing a boiler operator with a license lies in the reduced rigidity in the assignment of shifts and/or temporary sick leave or holidays.