Cooperation of PV farms with the transmission grid 3

[Mat_pe]

Hello everyone,

I was searching for some practical information about PV farms but didnt get it. So I got few questions:
1. As we know inverter adjusts voltage to grid level (or transformer does it). What happen when voltage in the grid is lower or higher? Is grid sending signal about voltage level to plant and by this we can change the ratio of step-up/down voltage?
Which situation is better? When we got higher voltage in grid or lower than nominal?

2. What protection divices we use in PV farm? What about their settings?
For example we got fault in grid and farm got disconnected - should it back on its own to the grid after passing the fault? How it looks like?
What can you say about short circuit power generated by PV plants? As we know this power is needed to ensure stability of the grid.

Thank you for reading my threat and any activity.

 

[cranky108]

The PV plant should see the transmission voltage and adjust. The transmission voltage is the signal, just like the transmission frequency is the signal.

PV plants on transmission should ride through a fault.
Many PV plants will reconnect after a time, and after being knocked off.

Do PV plants actually contribute into a fault? My experience is not very much, and no negative sequence.
Either they should act like a rotating machine, or provide modeling information to utilities.

Most PV equipment manufacturers really don't care about grid stability. They are only interested in making money.

 

[xnuke]

1. The output of a voltage-source inverter occurs through a series inductance. The voltage at the grid side of this inductance is monitored by the inverter, and the voltage magnitude on the inverter's side of the inductance can be varied by proper switching modulation.
[ul]
[li]If operated in unity or constant power factor mode, the inverter controls the magnitude of the voltage it generates on its side of the inductance. It matches the magnitude of the grid voltage it senses for unity pf output.[/li]
[li]If operated to produce a specific lagging or leading pf, it will produce a voltage magnitude that is lower than or higher than the grid voltage magnitude, respectively, to either consume or produce reactive power in proportion to its real power output.[/li]
[li]If operated to produce a specific reactive power, it will adjust its voltage magnitude with respect to the sensed grid voltage to do so.[/li]
[/ul]

2. Several protection functions are built into inverters to meet safety requirements per IEEE or IEC, but typically include over- and undervoltage, and over- and underfrequency, synchronization, anti-islanding, etc. However, external protection relays may be provided, and are sometimes required by the utility. Inverters will typically turn off their output when grid loss is detected after a period of time and will automatically resynchronize when the grid returns. Some inverters will inject/absorb reactive power during voltage sags/swells to aid the grid during such events with utility permission. Current output during faults is typiclly limited to be between 1.1 and 1.5 pu depending on inverter settings and capabilities.

ERPI has a good report on inverter functions available here.

xnuke
"Live and act within the limit of your knowledge and keep expanding it to the limit of your life." Ayn Rand, Atlas Shrugged.
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Gr8blu]

Mat: a properly connected source (which is what a solar farm really is supposed to be) monitors both sides of its transformer connection to the utility it serves. The source voltage (your side) is adjusted (by whatever means) to meet the measured voltage on the utility side (their side). Any difference between the measured voltage on the utility "grid" and where it taps to your transformer connection determines the direction of power flow. If the general "grid" setting is higher, power flows from the grid to you - which is undesirable, when you're trying to make money by supplying power.

The protective devices employed on the farm side of the connection are dependent on the location within the facility, and what - exactly - you're trying to protect. Typical sensing should at minimum include over/under voltage, over/under current, ground fault, and phase unbalance on the facility side of the connection. On the grid side you can measure the same things, although you don't really have to. All you need to know is whether the grid is able to accept power or not, which is accomplished by fairly simplistic voltage and current sensing. If the PV facility gets disconnected (from a fault on either side of the connection), control logic should be in place to determine when both sides are ready to run again, and to make the reconnection.

The amount of short circuit power from a PV facility is going to be limited by the maximum momentary output of the converter(s) that change the DC generated by the panel(s) into AC for distribution. Typically, this is limited to 150 % of the continuous rating, and is only available for 60 seconds or so (followed by sufficient "dwell" time to allow the converter components to cool back down). In contrast, a rotating supply can easily handle 3 to 5x rated output for a similar duration. You may also be limited in how fast your system can respond to this kind of demand.

Overall, your electrical system is only going to be as good as the weakest link. That may be the utility lines, the transformer, the facility internal wiring, the converter, or even a single PV panel. It might even be the control logic. In some cases, it might be the geography and weather for the farm's location itself.

Converting energy to motion for more than half a century

 

[xnuke]

Any difference between the measured voltage on the utility "grid" and where it taps to your transformer connection determines the direction of power flow. If the general "grid" setting is higher, power flows from the grid to you - which is undesirable, when you're trying to make money by supplying power.

Real power flow depends on the angular difference between the phasor voltages of the inverter and the grid across the output series inductance, not the magnitude difference. Reactive power flow depends on the magnitude difference.

xnuke
"Live and act within the limit of your knowledge and keep expanding it to the limit of your life." Ayn Rand, Atlas Shrugged.
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[stevenal]

If the general "grid" setting is higher, power flows from the grid to you - which is undesirable, when you're trying to make money by supplying power.

Since energy is conserved, I'm trying to understand this statement. How exactly does a PV farm motor? I'm imagining PV panels shining brightly into the night sky when the power flow reverses. Since this doesn't happen, I don't think PV farmers need to worry much about this "undesirable" effect.

 

[bacon4life]

Some very large solar farms receive a voltage/var control setpoint from the regional grid operator. In these very large installations, the control system of the solar farm can likely coordinate adjustments to individual inverters, mechanically switched capacitor banks, transformer taps, and/or power electronic compensation(SVC/STATCOM).

Smaller installations likely do not receive a control signal from the grid, and instead just react based on the voltage at the point of interconnection. An example of how this works for modern inverters per UL1741SA is on page 20 of

http://media3.ev-tv.me/UL-1741-Supplement-A_Webinar.pdf.

Legacy inverters often had no ability to provide voltage support.

Solar plants at full output can cause high voltage on distribution lines. Although somewhat counterintuitive, solar plants actually often work to lower the distribution voltage during peak production.

The link I posted also mentions various kinds of protection required of modern inverters. This is still an evolving field, so there are different sets of requirements throughout the world.

 

[Gr8blu]

xnuke You're correct. Phase difference for active (real) power, magnitude for reactive power. I'm too used to "oversimplifying" for a non-technical audience.

stevenal If the converter doesn't have the correct protection, a power flow from the grid to the converter will definitely cause some disruption. The panels themselves might well survive, though. I live in a fairly rural area - it's amazing what some people think is an okay approach, when the preferred method to tackle any project is DIY.

Converting energy to motion for more than half a century

 

[crshears]

Which situation is better? When we got higher voltage in grid or lower than nominal?

For the greatest operational flexibility, utilities definitely prefer operating their grids as close to the maximum voltage limit as comfortably possible.

For example we got fault in grid and farm got disconnected - should it back on its own to the grid after passing the fault? How it looks like?

The Independent Electrical System Operator [IESO} in my province does not permit automatic reconnection of ANY generation at the transmission voltage level; IESO concurrence with reconnection must always be obtained first.

Most PV equipment manufacturers really don't care about grid stability. They are only interested in making money.

I also find this to be true; almost across the board, PV operators prefer to operate their plants in a manner that holds the reactive power flow on the low side of their grid connection transformers at unity power factor, passing absolutely no more current through their transformer's low-voltage winding than necessary, as this maximizes transformer life and minimizes equipment degradation; they usually care not a whit what happens to the system voltage as long as they can remain connected.

They usually only agree to undertake to either provide or absorb lagging reactive power to the grid if it's negotiated right into the operating agreement well before the first shovelful of dirt is ever turned to build the plant, in other words, only if there's money in it for them.



CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[jrostron]

Is there much of an issue with congestion? That is power that can not be put into the grid because it would raise the local voltage too high?

 

[itsmoked]

That is power that can not be put into the grid because it would raise the local voltage too high?

Certainly! That's why residential grid-tie is all being forced to comply with UL1741 rules that include reducing output to prevent local distribution voltage rises. Some areas require a web connection to the utility so they can order your system off the grid. Some systems will not power up and supply anything to the grid until the utility's controller has authorized it. Witness the myriad Youtube videos of people who installed Tesla Power Walls only to discover they won't do anything until authorized by the utility's computer. In some cases the utilities don't even have a system in place to authorize operation so a powerwall is left useless for months. This can be really aggravating as during peak solar periods the utility can throttle back all the connected solar "to keep the voltage low" and then as soon as the solar availability drops off turn it all back up when no one has anything to offer. This can toast your ROI.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Mat_pe]

In last days i found idea called "PV-STATCOM". This solution uses rest of inventer's available power. For example: if inventers works on 30% of nominal we can use 70% of his nominal as a statcom.
I tried found some more information about it, mostly about how connections with grid looks like, but got only some articles.
Did you heard about it? Any inventer producer implemented that(i didnt find anything)? Are there also devices for PV farms not microinstallations?

 

[DBL-E]

I think one of the challenges in addressing these questions is that problem of 'it depends'. Mentioning PV farm and transmission grid, I envision something from 5 MW up to 400 MW+. These have very different impacts on the grid and have different requirements for their interconnection. xnuke mentioned the 'smart inverter' capabilities, equivalent to that 'PV-STATCOM' concept, and this operation is described in the epri document they linked previously, starting in the volt-var section, that gives a description of operation-

ERPI has a good report on inverter functions available here.

The specific settings again depends on the results of the interconnection study/agreement.

For a utility scale PV, you will have external protective relaying. NERC PRC requirements direct protection/coordination/ride-through characteristics for utility scale PV, but typically a utility has their standards available for general protection expectations. This PV studies webinar is pretty useful to visualize what engineers are evaluating for various stages of utility scale PV developments.

 

[jraef]

Using STATCOMs for PV solar installations is not new, I have seen separate STATCOM systems being advertised for that use for over a decade. I think the only “novel” aspect is that people making the utility scale grid-tied PV inverters are now incorporating the STATCOM functionality into the inverter system since much of the hardware is already there.


" We are all here on earth to help others; what on earth the others are here for I don't know." -- W. H. Auden