Network project Power to X Siemens Germesa/Institute of Technology.
Calculation of cooling water demand, heat exchanger and pumping technique for Power to X (hereafter referred to as PTX)
Under the project management of the Danish Technological Institute, Copenhagen Pump calculated the cooling water requirement for cooling the heat loss resulting from the electrolysis production of Hydrogen/Hydrogen (H2). We assessed MTBF Mean time between failures for a plant lifetime of 25 years and the hydraulic challenges of Open Loop and Closed Loop systems.
The calculations were carried out on a 500 MW and a 21 MW PTX load of offshore wind turbines.
Using the 3 sketches below, we explained the advantages and disadvantages. We gave recommendations for pumps and heat exchanger solutions, illustrated C02 foot print, energy losses, maintenance etc.
The two heat exchanger solutions that we assessed were a traditional plate heat exchanger (open loop) versus a submerged tube cooler (closed loop). The pumping solutions and hydraulic challenges derived from this, we have also considered.
The project has involved leading experts from manufacturers, from our private network, consulting engineering companies and universities abroad specializing in relevant physics.
We work from a model where we get an overview of the relevant physics, commercial needs and derive from this the technical challenges and technical solutions. We have a special focus on utilization and conversion of energy and maintenance.
Parts of the conclusion are presented below:
Conclusion:
Written by Thorbjørn Schrøder General Manger Copenhagen Pump
Chief engineer/EBA.
Pumps, pumping technology, hydraulic considerations, seawater cooler, fouling, maintenance, energy consumption.
Based on the assignment from the Danish Technological Institute 500 MW(h) PTX plant at Siemens Germesa/ Green Hydrogen, I have investigated different technologies and challenges based on open loop and closed loop illustrated by fig1, fig2 and fig3.
As no other technical data was available, I have assumed 2 basic load scenarios based on 500 MWH
A)500 MWh over 24 h
B)500 MW constant load.
Cooling water loss 20% of the power produced
South Sea delta P 12C
Cooling water/Glycol 30% delta T 60-40C
Geometric lifting height 20 m
Friction bar 1 bar
Estimated pump power about 1 MW corresponding to 1-2/1000
By ordinary calorimetric calculation Q= MxCxΔT, these ratios give 71.6 m3/h per 1 MW cooling capacity
500 MWh over 24 h (20.83 MW) gives a cooling water loss of 4.16 MW,
corresponding to cooling water consumption of 297 m3/h
500 MW gives a cooling water consumption of 7128 m3/h for 100 MW cooling water loss.
The pump power at 3 bar is about 1 MW, whereby closed loop can save 2/3
At an electricity price of DKK 1 and 8760 t/year, this corresponds to an operating cost of 1000DKK/hour or about 9 mil. and the savings will be about DKK 6 mil. The simple payback time is thus 23 years, however, without taking into account other costs, it will require a comprehensive technical analysis to properly assess the cost structure of the alternatives. However, it should be noted that a legal ban on Open Loop is not at all unlikely.
The cooling water consumption can be divided over decentralized plants where each turbine or a few turbines have their own independent cooling circuit of 297 m3/h or one can imagine a central cooling center/production platform of 7128 m3/h.
For both scenarios, you have to choose between two technologies A) Open Loop B) Closed Loop
For Open Loop fig 1, the kinetic energy cannot be easily and cheaply reversed and in practice this means that you have to pay it for the geometric lifting height in the example of 2o m.
The physical argument is found in the mechanical physics of pressure and heat theory where the vapor pressure of -1 bar is obtained at - 10 m, whereby the two columns of liquid lose each other. Water is able to absorb compressive forces but not tensile forces. This is a mistake that can be overlooked, as in open loop systems the head can also be subtracted for systems with less than 10 m delta H. This is how a siphon works.
For Closed Loop fig 2, the two geometric lifting heights cancel each other out and you only have to pay for the friction loss.
The physical argument is found in the law of conservation of energy.
It is theoretically possible to recover some of the kinetic energy at the bottom of the open loop with a turbine. From theory on the conversion of kinetic energy in wing-driven machines, it is known that theoretically a maximum of 60% of the kinetic energy can be converted and in practice for small machines much less, it is therefore not commercially feasible to attempt energy recapture as shown in figure 3. However, there is equipment for this purpose, typically propeller turbines.
Heat exchangers:
I have spoken with a manufacturer of plate heat exchangers Alfa Laval and the two cooling water scenarios are absolutely realizable and known technology from the oil industry and shipping respectively. Regarding maintenance and service life 25 years, annual inspection, disassembly and mechanical cleaning / high pressure cleaning must be expected 1-2 years.
Thermally the technology works well and is known technology. Fouling between the plates is not rated high by the manufacturer while in the inlet it will be more typical and it is recommended to install any automatic/sequential back flushing equipment.
My overall assessment and experience with the plate exchanger is that it is a solid solution.
Prizes and awards
4-5 MW 300 m3/h DKK 600.000
100 MW 7-10000 m3/h 6-8.000.000
Seawater pumps:
There are two options for installing seawater pumps.
There are advantages and disadvantages to both solutions. The lifetime of the submerged pump is typically somewhat shorter and not maintenance-free.
The traditional bronze ballast/fire/seawater pump typically lasts for 25 years with periodic maintenance on shaft seals.
The pumps are available up to about 2400 m3/h where 4 pumps can cover the 500 mw demand of 7-10.000 m3/h
300 m3/h pumps for 20 mw demand are completely standard ship pumps with 55 kw electric motors.
If you choose a central solution with 7-10,000 m3/h, my assessment is that a hull mounting* below the waterline, of a few large pumps, will provide the most problem-free pumping solution**, while the decentralized solution is probably more easily solved with 10-12" submersible pumps.
*mfb for wave height, construction heights and technical solution in general. Possible consideration of floating platform or ship, rather than standing platform. This solution may also be considered in combination with new clear and compression plants for high-pressure hydrogen, possibly for the transport sector and placed in connection with ports. Denmark has a tradition of shipbuilding, ship design of special ships and floating factories. Not dealt with further in this report.
**Examples of floating structures are shown in the publication offshore wind solution.
Sea Bed Cooler:
The Futures Submerged Sea Bed Cooler is a tubular cooler construction placed on the seabed. The system makes it possible to use the closed loop technology fig2, which saves a large part of the pumping work. You save the seawater pumps and can insert a good inline pump in simple metals, with a good shaft seal.
According to Future, the tube cooler is made of alloys that delay/prevent fouling and a 5 yearly high pressure cleaning of the bottom is sufficient (diving work). If you look at e.g. galvanized constructions, you will find a reduced fouling compared to e.g. untreated fiberglass or wood. After talking to Future it sounds plausible and they have proof of concept in Norway. According to Future, there is now a ban on the use of open loop systems in the German part of the North Sea. Overall, my assessment is that the technology makes very good sense. The offer obtained is a 5 MW cooler and costs approx. 10-12 million NKK which is a factor of 10 compared to the plate exchanger. In terms of energy, maintenance, it is my assessment that the advantages outweigh the disadvantages and probably also the construction costs, which can be carried out technically "simpler". I have not considered the price of the tube cooler other than that it is a factor 10 higher than the plate cooler, but the price probably depends on several factors that must be assessed in a more complex analysis on a large scale. Basically, the idea is good, as it has a lower co2 footfrint, does not require the emission of toxins and has a long operating time MTBM 5 years.
Fertilization in general:
It is my experience that fouling is very much related to materials and speed. For example, brass propellers that are sailed often, typically do not foul on the parts where the speed increases due to diameter increase, while at the center, you see few fouling. If the propeller is not sailed, the entire propeller is quickly fouled. There are chemical solutions in the form of copper-based paint, electrolysis of hypoclorite, zinc, etc. which are known chemical treatments.
According to Alfa laval, the flow velocities between the plates are so intense that fouling does not occur between the plates, whereas normal fouling must be foreseeable. Manufacturers of pasteurization plants for ballast water also report very long operating intervals of plate exchangers in their plants. This indicates that high heat has a favorable control effect.
Seawater systems on ships are typically not continuously chemically treated while underway. However, in some cases there is a dimensioned overcapacity of seawater of 300%. The seawater coolers are typically drawn 1-4 times per year depending on trade areas and high pressure cleaned.
Uncritical use of chemical treatment should probably be subject to closer analysis.
Cooling water pumps fresh water, possibly with glycol added:
We have 3 years of positive experience with Finnish Kolmeks pumps from brake systems in wind turbines. With this type of pumps we have not had any maintenance or operation related breakdowns and expect a MTBF factor of min 10 years. This is because the pumps are well balanced and use the best single mechanical shaft sealing technique. From land installations we also see very long intervals in cooling water systems. The pumps are inline pumps and thus have a very small foot print. The same pumps can also be used for pumping lye in connection with the electrolysis, alternatively, magnetically coupled pumps or Caned motor pumps are used, which are seal-free.
FW.Stainless steel cooling water pump 300m3/h x 55 kw amounts to DKK 100.000
FW.Cast iron cooling water pump 300m3/h x 55 kw amounts to DKK 55.000
Technical solutions: