Controlling the rotational speed of a generator

[Dawsonh4]

Hi,

I am working on a project that would insert a generator/expander system to a pressurized stream of gas in order to generate electricity.

I would like to control the rotational speed of the generator in order to adjust the gas stream flow rate.

Would it be possible to control the rotational speed of a generator through some sort of applied resistance? I am thinking something like an eddy brake.

Additionally, would large amount of resistance applied to the generator result in more power generated?

I appreciate anyone's input.

 

[anorlunda]

Welcome to PF.

To give a quality answer, we need more information.

AC? DC? What kind of generator?

What scale? 1 watt? Kw? Mw?

What is the load?

What are you trying to accomplish? Practical power generation? Just tinkering?

 

[Baluncore]

I would like to control the rotational speed of the generator in order to adjust the gas stream flow rate.

Are you trying to generate maximum energy from the flow, or to regulate the flow of gas?

An alternator will extract energy at a rate regulated by field current. That generated electrical energy must go somewhere.

The turbine design and gas flow will regulate the maximum unloaded RPM.

We need more information on the application.

 

[Dawsonh4]

Thanks for the reply.

Currently this is just a theory that I am trying to test the efficacy of. I have more of a chemical/mechanical background than electrical - I am out of my element.

It could be any kind of generator and likely a bigger one (300kw). The pressure drop would be in the 1,000s of psi and it would ideally be connected to the grid. The input pressure is variable and would be looking to fix the outlet pressure. A simple scenario would be an inlet pressure range of 1,000 to 2,000 psi and a fixed outlet pressure of 200 psi.

Number one priority here is regulate the flow of gas. Second priority is maximizing the energy generated.

I understand the turbine/expander would need to be designed in tandem with the generator. For this discussion I was hoping to focus on the generation side of things. The side I know least about.

Thanks for the much needed help!

 

[Baluncore]

Restating the problem.
A gas reservoir at high pressure (1000 to 2000 psi) provides the input. The outlet pressure will be regulated to 200 psi, with flow rate varying continuously with gas demand by the low pressure useage. Co-generated energy is transferred to the electrical grid.

There are a number of interlocked control loops that must be managed.

Ideally, the generator would spin at a fixed synchronous speed, with the energy flow regulated by field current. There are more complex variable speed generator solutions.

I expect the biggest problem you will have will be in adjusting the turbine for the range of demanded flow rate, over a variable input pressure.

 

[jrmichler]

You are describing a back pressure turbine (search the term). Many, if not most, paper mills use them. They generate steam at, for example, 1200 PSI, then reduce the steam pressure to the 50 to 75 PSI needed by the paper machine dryer section. A 300 KW back pressure turbine would be tiny by paper mill standards, where 5 MW would be more typical. I'm sure that other industries also use them.

A back pressure turbine installation needs to be engineered by a team with specific expertise in them. They need pressure relief valves, bypass valves, isolation valves, controls to synchronize to the line when starting up, and more.

Don't forget that the gas will come out COLD, so calculate the gas leaving temperature and figure out how you will deal with the cold gas.

 

[anorlunda]

If you connect it to the grid, you must comply with the technical requirements of the local power company. You will need a consulting engineering firm to design the whole thing.

 

[Dawsonh4]

Thanks.

A back pressure turbine is exactly what I am referring to. Thanks for honing that in for me.

I am still curious if there would be a way (without mechanical valves) to control the RPM of the generator and therefore slightly adjust the outlet pressure of the steam.

For example - if the outlet pressure needed to be reduced to 120 psi from 130 psi the turbine may apply more resistance to the stream by adding resistance to its rpm.

 

[Baluncore]

For example - if the outlet pressure needed to be reduced to 120 psi from 130 psi the turbine may apply more resistance to the stream by adding resistance to its rpm.

The game you must play is too complex for one human brain. You will need some tuned mathematical software. You cannot expect to juggle it through one variable.

For the output pressure to reduce, the flow through the turbine must be reduced, to below the LP flow demand. At the same time, that output pressure change will increase the pressure drop across the turbine. But without changing turbine geometry, the increased pressure drop will increase the flow, which is the opposite of what you really need.

If the generator is a synchronous alternator the turbine speed will not change with the load, it will remain locked to the grid frequency.

The software will need to continuously adjust both turbine geometry and generator excitation, while avoiding instability or oscillation.

Is the working fluid steam, or a hydrocarbon gas ?

 

[jrmichler]

Everything said above. There is more to connecting a generator to the grid than just flipping a switch.

This happened in the paper mill where I was working at the time: The operator switched a 5 MW generator online without properly phasing it first. Result: BOOM, then six months to repair.

 

[Dawsonh4]

There would definitely need to be a controller monitoring multiple things to make this work.

The flow may increase if you increase pressure drop. I am less concerned with flow than I am of pressure. It is a hydrocarbon stream.

Would an induction generator with a regenerative FVD be a practical solution?

 

[anorlunda]

I think you're approaching it backwards. Any electrical generator is merely a conversion device. It converts mechanical power produced by the turbine into electrical energy. Neglecting losses, the two powers are identical, thus conserving energy. Changing voltage, excitation, RPM, or any other thing does nothing to change conservation of energy.

So start upstream. How much power is available in the gas stream? How much of that is the turbine able to convert to mechanical power? Where does the rest of that energy go? Then how much of that is the generator able to convert to electric power. Last stage, how is that electrical power transported to the loads?

I'm saying that you can't practically design the process starting at the electrical end.

 

[Dawsonh4]

I think you're approaching it backwards.

I know this is definitely a backwards approach, but it's the reason I am asking for input. There are plenty of options for designing and installing a properly sized generator/expander system in a gas stream - see CAES.

My reason for working backwards is to avoid JT or control valves as much as possible. If it is possible to limit the amount of valves by giving the generator a larger operating range and adjusting its speeds to help control pressure then it might be possible to harness more energy.

 

[jrmichler]

There are plenty of options for designing and installing a properly sized generator/expander system in a gas stream

Since you already know this, then you should know that the best way to ...

harness more energy.

is to:

1) Get a quote for a properly sized generator/expander,
and then
2) Ask the supplier about options for increasing overall system efficiency.

They will likely give you more than one option. It will then be up to you to compare the capital cost, operating cost, and value generated by each option.

 

[anorlunda]

My reason for working backwards is to avoid JT or control valves as much as possible. If it is possible to limit the amount of valves by giving the generator a larger operating range and adjusting its speeds to help control pressure then it might be possible to harness more energy.

As @Baluncore already said, a synchronous generator runs at synchronous speed, regardless of the power output. Even an induction generator runs at close to synchronous speed (limited slip) through its whole power range.

Maximizing power while also regulating pressure requires analysis of the upstream process. In other words, the gas end. As a first approximation, the power delivered by the gas is proportional to mass flow times pressure. So how does gas production vary with pressure? If it is invariant with respect to pressure, then maximum power is achieved with maximum achievable pressure and pressure is not regulated.

Most steam boilers work the other way. The boiler pressure is relatively constant and the mass flow rate is regulated to change power. To make more mass flow without moving any steam valves, you burn more fuel in the furnace.

So you must start all the way upstream. How are gas production rate (mass flow) and pressure related in the process? In the OP you said "stream" rather than flow rate.

 

[Dawsonh4]

The issue I run into when reaching out to vendors/suppliers is that the sales teams is pushing their products rather than looking at things differently. They are not the R&D team.

I have reached out to them, received quotes and done the math on ROI. Everything checks out and looks like a good investment. However, I want to know if there is a way to make it better. All vendors I have gotten in touch with answer the control problem in the same way. They propose an intake throttle that passes excess flow stream around the expander.

I don't want to waste that additional pressure that is being routed around the system. This is why I am proposing a expander/generator system that can take advantage of variable RPM to help allow more or less flow through.

 

[Dawsonh4]

As @Baluncore
Maximizing power while also regulating pressure requires analysis of the upstream process. In other words, the gas end. As a first approximation, the power delivered by the gas is proportional to mass flow times pressure. So how does gas production vary with pressure? If it is invariant with respect to pressure, then maximum power is achieved with maximum achievable pressure and pressure is not regulated.

So you must start all the way upstream. How are gas production rate (mass flow) and pressure related in the process? In the OP you said "stream" rather than flow rate.

Upstream is a variable flow and pressure reservoir. It typically follows a production type curve for both flow and pressure. Pressure can start as high as 11,000 psi with flow around 14,000 MMSCFD. They then being to decline.

JT valves are put into place to control that pressure and flow. The focus is mainly on pressure as hardware further down stream can only handle a couple thousand psi max. The JT valve is constantly adjust as the input pressure changes.

 

[jrmichler]

Upstream is a variable flow and pressure reservoir. It typically follows a production type curve for both flow and pressure. Pressure can start as high as 11,000 psi with flow around 14,000 MMSCFD. They then being to decline.

I have reached out to them, received quotes and done the math on ROI.

A project like this starts with a matrix with flow on one axis, pressure on the other axis, and percentage of total flow (or time) in each cell. The supplier then provides kW generated and bypass flow rate for each cell in the matrix. I assume that you have already done this.

The next step is use that information to calculate the percentage of energy in the gas that is wasted through the bypass valve, then go back to the supplier(s) and ask about options to increase the total energy from the system. They will normally have options that generate more electricity, but at increased capital cost, and possibly increased maintenance cost.

Suppliers are all too familiar with customers that make purchase decisions by looking only at capital cost. Customers that are willing to pay more to get more need to specifically ask for it.

 

[Baluncore]

The issue I run into when reaching out to vendors/suppliers is that the sales teams is pushing their products rather than looking at things differently. They are not the R&D team.

Take the best system on offer now, with a warranty for a standard system.
That way you will reap the rewards over a longer period.

Perfection is the enemy of progress.

If you procrastinate for more than 5% of the depreciation time, trying to save another 5%, you will have higher costs. Once you have a functioning system you can do a cost and efficiency analysis while you are saving energy and establishing a benchmark. Stepwise refinement requires that you take the next step.

You dream of the first best, but perfection is unobtainable.
The second best will not be available until it is too late.
To win, you must install the third best today, and damn the imperfection.
Let the accountants decide between your summary of the immediately available systems.