← All episodes
S2E02

Preventing Europe's Next Blackout: Flexibility, AI & the Business of the Grid with Philip Hahn |S2E2

with Philip Hahn· Cyber Grid· 54m

TL;DR

Philip Hahn discusses grid flexibility, virtual power plants, and AI's role in stabilizing Europe's evolving energy system.

Synopsis
The episode features Philip Hahn of Cyber Grid, an expert in flexibility and virtual power plants (VPPs). He details his journey from EVN, where he fostered flexibility, to leading Cyber Grid, emphasizing the critical role of VPPs and automated trading in maintaining grid stability amidst the energy transition. The discussion covers the evolution of energy markets, the definition and application of flexibility as a service, and the challenges and opportunities in integrating diverse assets like hydro and batteries. Hahn also addresses the importance of cross-border balancing markets (Picasso, Mari) and cybersecurity in a digitalized grid. He highlights the rapid energy transition in Central and Eastern Europe and the need for clear regulatory frameworks, particularly Austria's new electricity law, to support investment in new, system-friendly assets. The conversation concludes with insights on AI's future role in grid operation and advice for new energy professionals.

Key metrics

by the numbers · 7
  • 2000-2001
    EU unbundling start
  • 11 years
    Market change duration
  • 96
    Daily energy auctions
  • 15,000 euros/MWh
    Max balancing energy price
  • 30 seconds
    FCR activation time
  • 70%
    Austrian hydro share
  • 10-15 years
    Electricity law update frequency

Topics

8 tags
Grid FlexibilityVirtual Power PlantsAncillary ServicesMarket AutomationCybersecurityEnergy PolicyBattery StorageCross-border Markets
Stats
Duration
53m
Words
8.1k
Questions
56

Timeline

10 chapters
  1. Guest Introduction and Background

    Philip Hahn from Cyber Grid is introduced as an expert in flexibility and virtual power plants, discussing his electrical engineering background and early career at EVN.

  2. Early Flexibility Projects at EVN

    Hahn describes his initial role at EVN, working on a flexibility management system project and gaining knowledge through both successes and failures.

  3. EVN's Strategic VPP Adoption

    Hahn explains EVN's decision to actively engage in flexibility markets, leading to the adoption of Cyber Grid's VPP solution for automated and scalable asset management.

  4. Market Automation and AI's Role

    The discussion covers the significant shift towards algorithmic trading and autonomous operation in energy markets, emphasizing the importance of human oversight for critical issues.

  5. Defining Flexibility as a Service

    Flexibility is defined as controllable generation or consumption assets, and Cyber Grid's 'flexibility as a service' helps asset owners monetize this capability on ancillary service markets.

  6. Optimizing Assets for Ancillary Services

    Hahn details how different asset types, like biogas plants, are optimized for various ancillary service markets (FCR, AFRR, MFRR) based on their activation speed and capabilities.

  7. VPPs, Grid Stability, and Cybersecurity

    Virtual Power Plants are presented as a 'last wall of defense' against blackouts, with increasing focus on cybersecurity in the context of a digitalized and interconnected grid.

  8. Cross-Border Balancing Markets

    Explanation of Picasso (aFRR) and Mari (mFRR) as cross-border balancing markets, using a firefighting analogy to illustrate the interplay of regional and international flexibility.

  9. Cyber-Knock Platform and Market Rules

    Details on Cyber-Knock, Cyber Grid's platform for automated bidding in capacity and energy markets, designed for scalability and efficient asset integration under harmonized European rules.

  10. Regulatory Challenges and New Electricity Law

    Discussion on Austria's new electricity law, highlighting the need for clear definitions of 'system-friendly' assets to provide legal certainty and encourage investment in new grid technologies.

Key insights

5 takeaways
  • 01

    Grid complexity exceeds human systems

    The energy grid is described as the most complex system humans have ever created, surpassing even space travel or the internet, making its operation inherently challenging.

  • 02

    Flexibility market needs liquidity

    Extreme price spikes in ancillary services (up to €15, 000/MWh) indicate a lack of market liquidity, highlighting the urgent need for more controllable flexibility to stabilize the grid.

  • 03

    Regulation lags technology adoption

    The energy sector's rapid technological advancements, especially in flexibility and digitalization, are outpacing regulatory frameworks, causing investment uncertainty and hindering faster grid evolution.

  • 04

    Human oversight critical for AI

    Despite increasing automation, human responsibility and oversight remain crucial in grid operations, especially for critical decisions during extreme scenarios like natural disasters, where AI may lack the full contextual picture.

  • 05

    CEE accelerates energy transition

    Central and Eastern European countries are making rapid progress in their energy transition, compressing decades of market liberalization and digitalization into significantly shorter timeframes, often leveraging existing assets and new battery technologies.

Pull quotes

3 quotes
  • Today I'm sitting here talking more or less about the same but before I was on the receiver side so Yeah. that was the time when when I was allowed to gain a lot of knowledge.
    Philip Hahn
  • I think our grandparents provided all of this and they thought long-term in decades and this is still our benefit today. But, now we see a huge fast energy transition.
    Philip Hahn
  • AI strength is to do this this autonomous processes and to autonomously fast secure work.
    Philip Hahn
Transcript1296 cuesClick a timestamp to jump
  1. Welcome to a new episode here in the
  2. energy bridge. Today we have the
  3. pleasure to have Philip Han from Cyber
  4. Grid. Expert in flexibility and virtual
  5. power plants. We're going to explore
  6. your interesting profile because you
  7. were part of the utility when they
  8. acquire
  9. virtual power plant and now you are kind
  10. of head of it. So
  11. interesting
  12. chief to be part of the both part of the
  13. table kind of in the same solution. So
  14. it's amazing. Thank you for being
  15. here today. Yes, thank you for having
  16. me. It's a pleasure for me. Yeah, thank
  17. you very much Philip. you
  18. were part of our last forum as well. You
  19. you gave a very interesting presentation
  20. and since then
  21. we have been chasing you but finally we
  22. made it. So very welcome. Okay, thank
  23. you.
  24. I hope I made it not too hard but I'm
  25. not so active on LinkedIn.
  26. So thank you that you had the patience.
  27. I'm sure we go for it. Yeah, Philip
  28. please
  29. walk us through your strong technical
  30. background and also your education
  31. journey and your work journey getting to
  32. the head of flexibility at Cyber Grid.
  33. Like how was the journey and what you
  34. did? Yeah, I never planned it. Business
  35. development but
  36. I was studying electrical engineering on
  37. the technical university. Then in my
  38. bachelor's I need to decide should I
  39. make my master's more into direction
  40. automation or energy.
  41. Back then I chose first automation but
  42. with an internship at EVN everything
  43. with EVN started back then and
  44. that
  45. woke up my interest in the energy
  46. sector because I was working in it.
  47. Very quickly I had some responsibilities
  48. in projects and it really was for me
  49. fascinating. Before I was super
  50. interested in it and then
  51. I was paid for my interest and I really
  52. enjoyed it and That's it.
  53. more and more on university side also
  54. picked some classes in that direction.
  55. very technical side first and then
  56. all the economy side of energy economy
  57. came then I would say naturally and that
  58. was also for me super interested because
  59. before it was just how the physics
  60. behind the working and all the
  61. electric magnetic field and so on and
  62. the generation is working and
  63. then immediately
  64. euros per megawatt hour and the market
  65. and also the all the
  66. the regulations behind it to create all
  67. of this that someone is willing to do
  68. and has the incentives. All of that came
  69. then
  70. naturally on the side and that was
  71. actually my my path from studies to
  72. work. And what was your exact role
  73. at EVN at first? At first I
  74. started with it was planned a one month
  75. internship.
  76. >>. One day before I started they asked
  77. me if I can stay two months because they
  78. need a little bit support in a project.
  79. >>. And actually it turned out that
  80. this project took then a little bit
  81. longer
  82. and still I'm more or less do the same
  83. so I started for actually a flexibility
  84. management system project. That was my
  85. my first day was I was sitting on a
  86. table with a huge project team and
  87. we were talking about how we can use
  88. which flexibility on different markets
  89. and so on. Today I'm sitting here
  90. talking more or less about the same but
  91. before I was on the receiver side so
  92. Yeah. that was the time when when I
  93. was allowed to gain a lot of knowledge.
  94. Actually most of the project also
  95. failed so by this by these mistakes and
  96. by these failing I think I learned even
  97. most and that was then later we will
  98. talk I think all of the further step
  99. and where then cyber grid Yeah. came
  100. into place and I was allowed them to
  101. lead the project on the EVN side. Yeah.
  102. So,
  103. before going to cyber grid, you were in
  104. EVN, which is a utility, and utilities
  105. are known for being
  106. a bit of conservative and focus on
  107. safety and reliance.
  108. And you were fostering flexibility from
  109. inside. So,
  110. what could you say there is the biggest
  111. misunderstanding about flexibility? The
  112. working environment of the old-fashioned
  113. utility since the separation this
  114. unbundling,
  115. I think the a huge transition already
  116. happened. And this transition we saw
  117. and when I came to have EVN, it was
  118. it was done. So, a liberal liberal
  119. market. So, to split distribution from
  120. generation
  121. and transmission.
  122. >> Exactly. Yeah, I think when it happened
  123. 2000 in the 2000s around 2000 2001
  124. something like this. It started. And
  125. then 2015 I came to EVN as a young
  126. intern. So, all of this happened
  127. already. And so, I in that
  128. department in that team, it was super
  129. innovative, super fast, a lot of
  130. interesting new topics. And before
  131. you were probably and the colleagues
  132. before were thinking maybe in decades
  133. and I think that was that was good and
  134. reasonable and that's the reason why
  135. we have a very
  136. good grid. We have a lot of hydropower
  137. plants in Austria and so, that's
  138. the work of the grand and grand
  139. grandfathers. So, my grandfather was
  140. this the shaking the concrete that it's
  141. strong. Really? Yeah, yeah, and he was
  142. for days not seeing sunlight and
  143. making all of these Danube hydropower
  144. plants.
  145. >> so you do grow up like with
  146. That was a long time before.
  147. But, your family is related with energy
  148. sector. actually, he
  149. >> He was good in shaking the
  150. concrete. He had no idea what he what it
  151. was good for. In the end,
  152. >> Who knew? but, I remember once
  153. together my my brother and I, we took
  154. our grandparents and then we checked
  155. these power plants and then he told the
  156. old stories when they drive from from
  157. the from the region they originally
  158. came. That's called Waldviertel wood
  159. quarter and they drove down
  160. then for spending a week. So, my
  161. uncle was working working there in this
  162. this company back then and
  163. they built I think one after the other
  164. like 10 of these hydro-power plants.
  165. Yeah, in the whole Danube, I think you
  166. have
  167. plenty of them. At least 10, I
  168. remember. Maybe more. Yeah, and you
  169. you have interesting
  170. Now, I came a little bit off topic. No,
  171. no. It's It's always nice to have those
  172. those kind of anecdotes. We had a
  173. cat. sorry. Now, I
  174. I remembered what I wanted to say
  175. before. So, I think
  176. our grandparents provided all of this
  177. and they thought long-term in decades
  178. and this is still our benefit today.
  179. But, now we see a huge fast energy
  180. transition. I think it's also good
  181. important and now
  182. on the other way, I try to be part of it
  183. and I think we all are and there
  184. flexibility,
  185. faster newer processes, new market, new
  186. incentives and so on. All all is
  187. happening at the same time.. So, so
  188. I didn't came to EVN to bring my
  189. knowledge and change everything and now
  190. we are fast because I'm here.
  191. They were already fast. They were
  192. already smart and I'm I was lucky to be
  193. part of that and also bring my
  194. strengths in it, but I was mainly on the
  195. receiving side, get all my knowledge
  196. about how this is working. They're very
  197. very good, fast, smart smart people
  198. there. And when you just arrived in 2015
  199. working for EVN evaluating VPP solution,
  200. then
  201. what was the motivation and what were
  202. the other options?
  203. And what did you see in the VPP market
  204. back then and now 10 years 11 years fast
  205. forward? Especially with the whole
  206. integration of all the new renewables as
  207. well. What changes do you see? Back then
  208. we had this decision to make, okay,
  209. this this previous project and the
  210. solution was not really turning out that
  211. way we wanted it to have. So two
  212. options, we stop all of this and say,
  213. okay, we market
  214. our our assets and our generation
  215. portfolio as we did always. So we don't
  216. change practices as as long or this the
  217. same way we always did. Or we kind of
  218. try to actively, that would be the
  219. passive way or to actively try to go
  220. into the flexibility market on several
  221. levels. So from the speaker utilities
  222. but down to the household customer. And
  223. then our decision was
  224. I think brave and I'm happy that we
  225. took the decision and then also it was a
  226. very good match with Cybergrid back
  227. then. And we made a market analysis, we
  228. tried to find out already had good
  229. experience what is what is not working,
  230. what is fully automated software was
  231. super important for us because we knew
  232. that's that's relevant when you have
  233. instead of
  234. one day ahead trade a day
  235. and then you have 96 energy auctions.
  236. And we knew 96 time a day nobody's
  237. willing to press every quarter hour
  238. button. So that was clear, fully
  239. automation was important, scalability
  240. was important, so we don't want to just
  241. have our own assets
  242. serviced, but also external assets. So,
  243. I think that was the active decision and
  244. I'm I'm happy we
  245. That's That's the only right decision
  246. decision to be active on the market. You
  247. see a chance and you take it and
  248. >> Yeah, and how do you see the market
  249. today?
  250. Comparing with the previous when you
  251. started. Yeah, when I started,
  252. I also had the luck and the chance to
  253. being a intraday optimizer, not
  254. full-time, but we have a trade
  255. trading desk and I was from time to
  256. time jumping into that
  257. into that role
  258. to be the trader in the night on the
  259. weekend or so on. So, all of these
  260. trainings, all of the system, all of the
  261. markets I
  262. actively participated in these last 11
  263. years, a lot changed. So, in the
  264. beginning it was not just the room and
  265. and the infrastructure changed, but
  266. also
  267. especially the degree of automatization
  268. completely changed. So, algorithmic
  269. trading
  270. more and more is automatized. Overnight,
  271. back then we need to observe it or and
  272. and be on the desk. Now,
  273. mostly can be fully automatized and
  274. this running autonomously.
  275. And just if there is then a issue
  276. and this is the most important
  277. thing to have limits, guidelines, and
  278. and good failure and notification system
  279. that
  280. the human strength is to have the
  281. responsibility. And
  282. AI strength is to do this this
  283. autonomous processes and to autonomously
  284. fast secure work. And also, Philip,
  285. before we dive in
  286. deeper,
  287. I want you to also tell us define
  288. flexibility and what exactly flexibility
  289. as a service means. Yeah, for us we are
  290. always talking about flexibility and
  291. I think most people are wondering what
  292. is this flexibility, yeah. And we
  293. see especially when we talk to customers
  294. or industrial customers, they talk
  295. with their language to us and they say
  296. we need this process heat there and
  297. that's why on the side there is this
  298. electricity generation and so on and we
  299. talk about the flexibility and then we
  300. need or we try also to be the translator
  301. to exchange in this between these two
  302. worlds and flexibility I see
  303. generation or consumption asset.
  304. have is controllable and have the
  305. possibility to increase
  306. the generation or the consumption.
  307. and have this flexibility within limits
  308. and it's planable controllable
  309. then then it's flexibility which can be
  310. used then for
  311. different services. We use it mainly for
  312. stabilizing the or to help
  313. the Austrian TSO to stabilize the
  314. frequency. but it's also for
  315. curtailments for instance or if there is
  316. a if the grid is overloaded then you
  317. this curtailment reserve stabilizing
  318. reserve it's called. If too much
  319. electricity is flowing then you need to
  320. reduce the electricity so there's two
  321. different things. One is to save the
  322. frequency that the generation
  323. consumption is in balance and the other
  324. one is that the grid is not
  325. overheated. And you managing
  326. flexibility as a service so who is the
  327. consumer and who is the provider? In our
  328. flexibility as a service team we
  329. have a big stream is our main customers
  330. are someone who has flexibility and want
  331. to have a revenue out of this
  332. flexibility.
  333. >>. So, whenever
  334. someone with a small power plant or a
  335. big power plant wanted to go on the on
  336. this ancillary service markets, then we
  337. are the perfect service provider for
  338. that.
  339. we call it customer onboarding. So,
  340. the customer comes to us and we guide
  341. him through all the process from
  342. the first sales contact to then
  343. contracting, pre-qualification, then the
  344. tele meter control connection to the
  345. asset, and then we also make a
  346. signal test, activation test, and then
  347. it's possible to earn money with it and
  348. have it have it on the market. That's
  349. that's just one stream and we have also
  350. some some very good running side streams
  351. like consulting for instance. So, we are
  352. also experts on simulating, calculating
  353. revenues for battery energy storage
  354. system. So, for best systems. Very
  355. demanded now. In very demanded, yeah.
  356. Interestingly,
  357. they are very specific usually. So,
  358. most of the people come and say, "Hey,
  359. just tell us what we can earn." We say,
  360. "Okay, you want to know it exactly or
  361. you want to know an average?" And this
  362. average is standalone somewhere
  363. then you can say, "Oh, it's earning that
  364. much." But usually you have grid
  365. constraints. You have a co-location
  366. maybe with a renewable energy generation
  367. so on
  368. and so on and so on. And in the end
  369. with all of that constraints,
  370. they still want to know, "Okay, if I now
  371. invest in that, how much money I can
  372. make that then?" And I think we are
  373. expert also in that. So, since 2 years
  374. we are providing that service and it's
  375. it's running excellent. And
  376. being part of within a utility, how much
  377. of that is an enabler and how much of
  378. that restrict you? on the utility
  379. side
  380. it's it's just
  381. enabling. The restrictions come more
  382. from regulation side or when the
  383. grid has not more capacity. But actually
  384. it's also an interesting conversation.
  385. Maybe we'll talk later then about the
  386. new
  387. electricity law the electricity law
  388. is that what is now supporting the
  389. grid? What is system supporting these
  390. terms and what is then kind of
  391. need to pay this this grid fees or where
  392. reduced grid fees and so on. If you
  393. could walk us through
  394. how do you
  395. manage to have a dump asset like a
  396. biogas power plant for example to a high
  397. value market asset?
  398. How do you do it?
  399. >> So
  400. we never call a asset dump.
  401. I know that that was called cool in the
  402. first place. But I understand what you
  403. mean. I think every every technology,
  404. every asset type has strengths and
  405. weaknesses. We can just use what what
  406. the asset is capable of. So there are
  407. restrictions in I don't know ramping
  408. time. Is it fast enough maybe for AFRR
  409. then
  410. we try to bring with good concepts
  411. and algorithm the asset to the higher
  412. market. I already mentioned some some
  413. terms. So AFRR is this secondary control
  414. or MF RR is the tertiary control. A
  415. little bit slower, a little bit less
  416. activations. And those are ancillary
  417. services at the end. Exactly. And you
  418. can provide it because you integrate
  419. more technology on it. Okay. Now I try
  420. to explain it briefly what is ancillary
  421. services. There are these three markets
  422. primary, secondary, tertiary.. In
  423. English, it's called FCR for frequency
  424. curtailment reserve. Yeah. AFR and
  425. MFR. AFR is standing for automated
  426. frequency restoration reserve, and
  427. MFR manual.
  428. And they have different ramping times or
  429. different time spans when they need
  430. to be activated. So, first is primary,
  431. that's immediately frequency
  432. measurement. So, it's measuring the
  433. frequency, and when I have There is a
  434. deviation, then it's it's activated
  435. directly on the asset. So, that's a
  436. little bit different to AFR. AFR is also
  437. kind of measuring the or APG is
  438. measuring the frequency within the
  439. control area of Austria and wherever
  440. because the frequency is a global.
  441. APG is checking the frequency
  442. deviation within the control area of
  443. Austria. Whenever activation in
  444. Austria is then necessary, they activate
  445. an AFR. And a little bit time delayed,
  446. so it's 30 seconds, 5 minutes, 12.5
  447. minutes for the activations.
  448. And that means how how smart is
  449. your asset to name it the other way,
  450. or how fast is the asset you can play
  451. on different markets. But So, is that
  452. time thing?
  453. It's mainly mainly a time thing. And
  454. what are the ranges?
  455. Milliseconds, seconds?
  456. >> That is These are the 30 seconds for FCR
  457. from
  458. till full activation time, this 5
  459. minutes. We can also say 300 seconds or
  460. this 12.5
  461. minutes.
  462. Yeah, and just
  463. getting you back from a bit of
  464. just we to also understand virtual power
  465. plants, and if you think that VPP
  466. it's also
  467. kind of truly a firewall against the
  468. blackouts that we see and in terms of
  469. grid stability because flexibility as a
  470. service is something that
  471. it also challenges stability and grid
  472. stability is
  473. the talk of the town what happened in
  474. Poland and also in Germany and last
  475. year in Spain. So, how do you see and
  476. also do you see VPP being part of the a
  477. firewall for all the stabilization that
  478. we
  479. I would maybe not call it firewall but
  480. maybe
  481. last wall of defense
  482. to hold the frequency and the frequency
  483. is a term to see that the if the
  484. frequency drops too strong or increases
  485. too fast but mainly drops too fast then
  486. then all of this blackout dangers
  487. happening and usually all this
  488. if you read then in the news the almost
  489. a blackout and so on if you go then
  490. through the
  491. days weeks later then a chronology is
  492. published and what exactly happened
  493. and most of the time it's a chain on in
  494. in failures mainly then
  495. a human failure somewhere happened some
  496. some operator maybe operated one
  497. important transformator wrong in the
  498. wrong direction and so on and these are
  499. then the most relevant of failures
  500. of this long failure chain but usually
  501. it's so complex the grid and the
  502. operation of the grid that once I heard
  503. it's the most complex system humans
  504. created right and humans were on the
  505. moon and we have the internet and
  506. and still the energy system is the most
  507. complex system of all of them and I
  508. think there's so many
  509. or already good implemented
  510. and secure borders that nothing is
  511. happening and actually in Europe
  512. we have very very less of this energy
  513. outages, right? so it's it's very
  514. well
  515. and strong and resiliently built
  516. their whole energy system.
  517. but still with a mix and chain of
  518. failures
  519. whenever there is a chance of failure
  520. and
  521. in thousand code lines there are four
  522. four
  523. code lines for sure wrong.
  524. so more and more of this complexity can
  525. also lead then to
  526. failures and if there is the right chain
  527. of failures then of course this can
  528. always happen. And with all of these
  529. virtual power plants connected and
  530. helping the grid to be the last line of
  531. defense, how are you addressing the
  532. potential cyber risks? Yes, I think that
  533. becomes also more relevant. Back then it
  534. was hosted in some server center at the
  535. utility and now more and more is
  536. outsourced somewhere in the cloud in a
  537. data center. It's leading in two
  538. different directions. On the one hand
  539. it's good that it's when it's in the
  540. cloud you cannot directly attack a cloud
  541. because usually it's part here, part
  542. there and it's back up in each
  543. other. On the other hand it's a lot more
  544. digitalized. So whenever a hacker is
  545. good then he cannot hack just one
  546. utility but if he's really good then he
  547. can
  548. hack the whole system in worst case.
  549. But it's always
  550. like increasing
  551. the hackers become better and the
  552. defense system becoming better all the
  553. time and I think
  554. it also helps that it's super
  555. diverse, that it's not just one system
  556. for a centralized for everything but
  557. there are different VPPs running with
  558. very intelligent IT sec people who who
  559. takes care and try to defend it all of
  560. this encryption, VPNs. There are so many
  561. mechanism to protect it to outside, to
  562. have a gear redundant, and so on and so
  563. on. So,
  564. in the end I think also hackers
  565. have a harder and harder job to Yeah.
  566. Actually, we're exploring the
  567. organization of hackathons, which is a
  568. IT concept, but to bring it into energy.
  569. So, we are introducing more and more the
  570. concept of energy security.
  571. And it could be cybersecurity, like
  572. these risks, or also infrastructure
  573. security. So,
  574. we can explore it later, but now I
  575. wanted to ask you for we have heard
  576. that in Europe is being used Picasso and
  577. Mari. So, if you could explain what they
  578. they mean and how
  579. how far the flexibility can goes in like
  580. one asset here
  581. in Austria can help the in other
  582. countries. You used to know how
  583. what do they mean and what how they
  584. work. Picasso is for aFRR and Mari, the
  585. M I remember it's standing for yeah,
  586. mFRR tertiary.
  587. And it's cross-border balancing markets.
  588. So, that means that you can trade
  589. your flexibility or your flexibility is
  590. used not in your own control area only,
  591. but it's also in can be also used
  592. then for others. Very important,
  593. there are borders and there
  594. grid constraints on the borders. So,
  595. just as much as the grid is able then to
  596. provide, for instance, in if an Austrian
  597. asset one of our asset is jumping in
  598. for a for Portugal
  599. deviation in the in the frequency,
  600. then of course, there need to be
  601. the capacity of the grid over the
  602. borders free that we can deliver it. And
  603. so there is still a core value of
  604. flexibility with within the own country.
  605. Well, for me it helps to explain it
  606. usually to compare the these services,
  607. these ancillary services with
  608. firefighting. I always compare this
  609. firefighting example. So
  610. that means in that term that within your
  611. region, within Austria, we have
  612. firefighters that can drive to Portugal
  613. to
  614. fight the fire. But at least a certain
  615. amount of firefighter need to stay in
  616. Austria
  617. to make sure
  618. they can also come then
  619. firefighters from Czech Republic to help
  620. in Austria.
  621. But still also
  622. the roads need to be free. It doesn't
  623. if the road doesn't provide the fire
  624. or cannot bring the firefighters then to
  625. to Portugal then they are useless,
  626. right? We also need to take take care of
  627. that. But that is the idea behind. And
  628. why I compare it with firefighters I
  629. before wanted to explain a little bit
  630. what is this FCR, EFR, MFR. They are in
  631. different markets. So one one important
  632. difference is the activation speed.
  633. One is also the likelihood of
  634. activation. So the activation
  635. probability is different in this in this
  636. market. And
  637. for instance FCR is in both
  638. directions is symmetric. EFR and MFR
  639. they are divided into positive negative.
  640. And MFR and EFR are divided into
  641. this
  642. capacity
  643. market and energy market. Power and
  644. energy you can also say it like this.
  645. And I like to compare it with
  646. firefighters because here it's clear for
  647. that service even if there is no fire
  648. by waiting you get paid. It's a good
  649. analogy. is capacity. And when there is
  650. fire and you drive with your firefighter
  651. car there and
  652. They drop water. So, that's the
  653. >> And then then there is a price for the
  654. water you
  655. you provide. It's a good analogy. Good
  656. Yeah. And
  657. just going deep into this
  658. firefighting machine that you have and
  659. you call it Cyber-Knock.
  660. Tell us about Cyber-Knock a bit as a as
  661. a
  662. the first as a platform.
  663. And also utility of it going
  664. forward compared to
  665. an operator.
  666. What will be the difference because
  667. how does the platform itself manage the
  668. wallet volatility of the market and the
  669. negative price or sequences like
  670. that? Now, I jump again a little bit
  671. back into time. Back then I was project
  672. manager to implement the Cyber-Knock at
  673. EVN, right? That's the new software.
  674. We're going to use it. And
  675. in that time also the Mari or first
  676. the Picasso go-live happened and then
  677. the Mari go-live.. And for us that
  678. was
  679. one of our main decisions was we now
  680. need a autonomous system
  681. which allows our assets and our
  682. firefighter cars, to come back to that,
  683. to participate active on the two
  684. markets..
  685. >> Capacity market is
  686. this day-ahead auction market.
  687. Austrian TSO make sure every day that on
  688. the next day there are enough
  689. firefighters..
  690. So, for that we need a fully automated
  691. system, but even more for the for the
  692. energy market because the energy market
  693. also gave them the chance, if you're not
  694. accepted with your bids on the capacity
  695. market, you still have the chance to go
  696. on the energy market and without getting
  697. getting paid for capacity you can get
  698. paid for energy and that's happening
  699. every quarter hour so 96 times per day.
  700. We kind of implemented
  701. all of this
  702. because of money rules
  703. back then when it went live and that
  704. was
  705. one of the biggest strengths of back
  706. or is still one of our biggest strengths
  707. but that was kind of very very new back
  708. then because
  709. we were also I think one of the first or
  710. even the first who were able to be
  711. active then on the because and the
  712. money market.
  713. Another very important thing for us was
  714. before we had the pain that it was very
  715. difficult when a new asset needed to
  716. be connected and it was a very pain.
  717. Every asset we always wanted to
  718. standardize it but every asset was
  719. somehow specific.
  720. And within a very stiff and
  721. old-fashioned software it was almost
  722. impossible to customer friendly. We
  723. always need
  724. also the software provider to also
  725. parameterize and it was
  726. very expensive, time-consuming, and a
  727. lot pain to add one external asset. And
  728. for us it was super important doesn't
  729. matter external or internal whenever
  730. there's a new asset we want to be fast,
  731. we want to be we want to do it by our
  732. own, we want to be scalable that we have
  733. the potential to bring a lot of
  734. flexibility into our VPP.
  735. And I think that was one of the biggest
  736. strengths but of course that that
  737. happened a lot a lot development that
  738. was our luck or our profit that EVN back
  739. then when we when we kind of
  740. once we called it we married a
  741. marriage EVN and CyberGrid that was one
  742. of our big strengths to develop then
  743. something together.
  744. To develop something for a European
  745. market and the Picasso rules are
  746. very similar. Doesn't matter if it's
  747. Germany, Austria, or Portugal,
  748. Croatia, Norway, it's very similar then.
  749. And that was if you have one system for
  750. for Picasso, then you can use it most
  751. everywhere. Okay, so Picasso Mari gives
  752. like the
  753. framework where your algorithm can play
  754. kind of to make bids into the market.
  755. Right? Yes, so Picasso and Mari are
  756. these cross-border markets and they have
  757. certain rules. For instance, one of the
  758. harmonized rule is capacity bid is 4
  759. hours long. And then our algorithm, of
  760. course, need to
  761. forecast,
  762. calculate, and all of this
  763. and send automated bids then to the
  764. trading platform and so on for this
  765. market rules. Okay, so the question is
  766. if is an algorithm and now we are using
  767. more AI, is everything automatized? What
  768. happen with a bid it's 3, 000 euros per
  769. megawatt hour? Do you still have this
  770. panic bottom? Where is the human
  771. control in the loop?
  772. Yeah,
  773. ancillary services and balancing
  774. energy is even more extreme.
  775. So,
  776. we have prices up to 15, 000 euros
  777. So, positive, negative also also that
  778. happened, but
  779. for me it shows that more flexibility is
  780. needed. That the market is not
  781. completely liquid yet.. it's it's
  782. really needed and sometimes we need to
  783. activate even very very expensive
  784. flexibility
  785. to be stabilized and it's a lot still a
  786. lot cheaper than a blackout..
  787. Yeah. In the end. And so,
  788. for sure, like on this side,
  789. we and our algorithm need to take care
  790. that we are not
  791. how is this called? I think in algo
  792. trading is this
  793. thick finger or so, that you
  794. place a bid, you mean maybe minus 15,
  795. and then it's minus 15, 000, or you
  796. compare you
  797. you want you want to type in megawatt,
  798. and then it's kilowatt or whatever. So,
  799. this this thick but this you need to
  800. take care, or
  801. the system need to provide you then a
  802. alarm if you really want to
  803. place this bid if you want to place it
  804. manually.
  805. And this parameterization, this
  806. strategy, this also this backup
  807. mechanisms and so on, they need to be
  808. trained, strengthened, tested all the
  809. time that this is not happen happening.
  810. In the end, there need always be a human
  811. in responsibility.
  812. So, it's maybe interesting similar
  813. thought is in autonomously driving, it's
  814. also the question who is responsible
  815. when a person gets killed, and
  816. who decides should be a kid killed or
  817. eight old people walking about
  818. on the
  819. on the street and so on. And
  820. and I think in our algorithms, we are
  821. not there yet to have this
  822. fully AI artificial intelligence
  823. yet, but very strong trained within
  824. borders
  825. system to be also built not for the
  826. perfect day, but for a day where there
  827. are some data missing, and
  828. where the previous system has a
  829. failure, and the
  830. interfaces is changing, and still
  831. you want to
  832. participate on the market, and
  833. it's it's quite an important
  834. service. Like that's why I like also to
  835. compare it with firefighting. And in the
  836. dystopian future,
  837. could you imagine, electrical system
  838. operated fully automated with AI,
  839. for example? It I'm not sure if if I'm
  840. more afraid or excited about that.
  841. >> Yeah.
  842. But I think,
  843. I'm more on the con- conservative side
  844. on that. I think the final decision
  845. should should be made by by human and I
  846. think we still have some advantages
  847. against
  848. above AI. And I think we have a
  849. still overall picture and we can
  850. take take more parameters into
  851. account. For instance, there was the
  852. flooding lower in lower Austria. Yeah, I
  853. was stuck, so I know it very vividly.
  854. Okay. And, in these times, a
  855. lot a lot of good human, decisions
  856. were made that to prevent even a lot
  857. more damage and even,
  858. risk for someone losing their
  859. life and so on. I'm not sure if in that
  860. extreme scenarios, AI is able to take
  861. the right decision and to always provide
  862. the full picture and all of the
  863. information. I think I don't know., I
  864. think in some ways, it could also be
  865. the other way around. In lower Austria
  866. with EVN, we were a little bit closer
  867. and saw all all what happened and
  868. all was all what was not happening
  869. because
  870. some very good decision were made up
  871. front.
  872. but in that scenario, I'm not
  873. sure. I would be still a little bit
  874. scared, when AI is taking all of this
  875. decision. I don't know. Yeah. Would you
  876. like someone at AI AI doctors operating
  877. you already?
  878. Yeah, that's maybe maybe with support,
  879. right? But there should be still a
  880. person
  881. standing next to it and observe it and
  882. and have the responsibility in the
  883. end. Yeah. Also talking about the
  884. traditional assets now. Yeah,
  885. since
  886. biogas and hydro they also flexibility
  887. options, right? So, do you think
  888. hydro in Austria there's a lot of
  889. hydro storage and you think they're as
  890. effective as going forward as Tesla mega
  891. back
  892. or how do you rate them as a flexibility
  893. asset?,
  894. what I like so much about hydro is I
  895. think as Austrians you need to like
  896. hydro. It's renewable, right? It's
  897. already here.
  898. >> And unlike nuclear. And it's maybe
  899. seasonal storage and we don't
  900. have nuclear and I think that's that's
  901. good as as it is. 70% hydro in Austria?
  902. I think the electrical energy is 70,
  903. yeah, two two thirds. Yeah. Around. And
  904. that's a lot. It's because we have the
  905. the Alps and we don't have that huge
  906. consumption in electricity.,
  907. still, but for
  908. 8 9 million and not for 80. What I like
  909. is about you have you can also have
  910. seasonal storage.. And compared to a
  911. battery
  912. a battery with the losses is losing
  913. energy anyway because it need to be
  914. cooled when it's operated and round trip
  915. efficiency is always less than 100% of
  916. course. There are some losses.
  917. Compared to a hydropower plants,
  918. whenever it's raining, it's it's loading
  919. your your storage, right? And it's built
  920. for decades. That that's why I really
  921. like hydropower plants. But
  922. nevertheless, it's impressive how
  923. how cheap batteries already are. It's
  924. definitely part of the solution. They're
  925. super fast. They're precise.
  926. And it they're definitely a very
  927. important enabler for our energy
  928. transition. Both technologies have have
  929. their strengths or to name also biogas.
  930. whenever we can use
  931. biogas generation to reduce the
  932. generation and store the biogas then for
  933. for later it's it's it's even the best.
  934. Then you can
  935. save energy. You can have it then later
  936. and
  937. then then a very simple flexibility
  938. asset
  939. has has a has a good impact to
  940. the overall system. Yeah. And now moving
  941. to Central and Eastern Europe
  942. region
  943. because Cybergrid was founded in
  944. Slovenia. You still have
  945. teams and projects there but you
  946. also were telling us that you have
  947. projects in other countries in
  948. Central and Eastern Europe.
  949. How do you think the challenge to
  950. implement more
  951. virtual power plants there it's a
  952. more related to the grid outdated
  953. infrastructure or more software
  954. skills
  955. like digital skills?
  956. Yeah.
  957. we as we as Cybergrid we have the
  958. Cyber NOC. We operated at EVN. we
  959. provided service.
  960. But also we have a software as a service
  961. stream. So we provide the software to
  962. other utilities. We have it up and
  963. running in Croatia, Bulgaria, North
  964. Macedonia and now we go also on
  965. the German market
  966. very soon. Maybe when the podcast is
  967. online, we already on the German market.
  968. so, we go to more and more and
  969. then a lot of other countries. And
  970. my my impression I'm not directly
  971. into that project, but my impression is
  972. that we are a little bit ahead because
  973. our our Mari go live and our Picasso go
  974. live, it was
  975. I think 2022, 2023, something like this.
  976. For we already had collaborations
  977. with
  978. with German TSOs,
  979. with Czech TSOs, and they were there
  980. was already a I would call it a pilot of
  981. of Mari up and running.. And our
  982. liberal- liberalization of the market
  983. happened longer time ago.. So, what
  984. I think what Eastern Europe is now doing
  985. is to make the same transition in a lot
  986. less time.. I think the
  987. what I see, they do a very good job.
  988. I'm always impressed how smart the
  989. people are who are leading the other
  990. project and the implementing all of this
  991. software solution. They are super
  992. curious, they are super interested in
  993. that, but all already bring so much
  994. knowledge. They also have very very good
  995. technical education and so on. So, I
  996. think they are able to do it. And
  997. I'm talking here
  998. they have everything in the future. I
  999. think they already implemented a lot,
  1000. but all of this cross-border is working
  1001. specially good if you have a lot of
  1002. borders where other market participants
  1003. are.. And I think that's a little
  1004. bit a bottleneck right now that for
  1005. instance when one country is the first
  1006. active member, then
  1007. they can cross-border trade, but
  1008. there is no country ready on the
  1009. other side of the border yet. But now
  1010. more and more countries are active
  1011. and we deliver also as things
  1012. software our cyber knock..
  1013. And that brings them now
  1014. more and more into into new markets in
  1015. cross border and I think they are a
  1016. little bit behind but they are
  1017. really running fast to come on the on
  1018. the same level. And I think in that
  1019. case digitalization is super important.
  1020. I think that's the first step to make
  1021. the
  1022. digitalization and then bring new assets
  1023. battery energy storage systems and so on
  1024. in a in a second step on the market or
  1025. even better would be in parallel if
  1026. you don't have to choose. If you had the
  1027. budget.
  1028. we heard in Bulgaria they're
  1029. just going
  1030. full throttle about batteries.
  1031. They're putting a lot of batteries. So
  1032. you think the
  1033. the digitalization is the low hanging
  1034. fruit and then going for the batteries.
  1035. Yeah or
  1036. to
  1037. to use use what you have already..
  1038. So when you have they also have a
  1039. hydropower plants we know because heart
  1040. of EVN group a utility in Bulgaria
  1041. and North Macedonia.
  1042. And all the all the assets they
  1043. already have or they also build new
  1044. assets and also they have a lot of PV
  1045. generation for instance and even a
  1046. lot more PV hours than was sun hours
  1047. than we have in Austria. So they have
  1048. other chances they should use it and
  1049. whatever they have they should already
  1050. bring them in into into a market and
  1051. make the market more liquid and when the
  1052. market is functioning and the
  1053. neighboring market is functioning and
  1054. the It's like a snowball. Then it I
  1055. think this is a snowball effect. And so
  1056. use what you have
  1057. and you need some digitalization to make
  1058. use of this flexibility and
  1059. best case at the same time build
  1060. new new
  1061. technologies and into the grid. Yeah,
  1062. and now going into
  1063. one of our thesis in the energy bridge
  1064. is the technology is going faster than
  1065. regulation and energy super regulated
  1066. sector.
  1067. And now we Austria has a new electricity
  1068. law.
  1069. So, if you could tell us briefly
  1070. what is it about and what would you add
  1071. to this law to enable more flexibility?
  1072. Yeah, I think every I don't know 10 15
  1073. years there's a new electricity law
  1074. coming as I think it's more supposed to
  1075. every 10 years or so when it feels
  1076. necessary. Maybe this also will be
  1077. shorter terms in future.
  1078. but I think all the all the energy
  1079. market was desperately waiting to
  1080. that it to be published. They needed to
  1081. plan. They need security.
  1082. they need legal certainty. Yeah, I'm
  1083. not expert on that, but we have still
  1084. some issue as long as that's not decided
  1085. and I still when now we recorded it's
  1086. not 100% decided.
  1087. and for us one of these terms is this
  1088. if an asset is system friendly, it has
  1089. some advantages. And what does system
  1090. friendly mean? Especially if you want to
  1091. build new assets. So,
  1092. when a hydropower plant is 80 years
  1093. old, then it's not waiting for the new
  1094. law. But especially if you want to build
  1095. tomorrow something new, you want to have
  1096. some investments into generation, into
  1097. also in our system to for decision make
  1098. decision making, we need clarity. And I
  1099. think
  1100. now part of this clarity we have,
  1101. but still not some certainty how it's
  1102. interpreted in some terms. For instance,
  1103. this
  1104. is this as is a battery system
  1105. friendly? Is it system friendly if
  1106. it's a co-location with a renewable
  1107. energy generation? And this need to
  1108. be decided better soon that all these
  1109. investments
  1110. keep on going and happening that we have
  1111. in this asset and this flexibility. So,
  1112. you would add more clarity in
  1113. definition. Yeah, I would
  1114. Two things, I think it helps if it's
  1115. clear up front.. If not, then it
  1116. need to be clear how it's interpreted or
  1117. how it's operated in reality. How a
  1118. control defines this is system-friendly,
  1119. that it can have a great fear reduction
  1120. and that's not system-friendly. It need
  1121. to be clear for all market participants.
  1122. Yeah, we had a
  1123. interview with
  1124. Wind Association.
  1125. And Florian told us that they prefer to
  1126. have clear rules even if they are not
  1127. completely aligned with them, but it's
  1128. preferable
  1129. to have rules. Clear.
  1130. That's I would agree on that and I think
  1131. You have maybe less or different
  1132. investments or whatever, but as long
  1133. it's not decided, nobody's investing.
  1134. Everybody's waiting what's going to
  1135. happen. When it is sure, let's wait. And
  1136. and I think that's
  1137. helping nobody. Yeah, Philip, looking at
  1138. the time
  1139. we
  1140. we it's very insightful, but using the
  1141. flexibility to get into our
  1142. last round.
  1143. The time flexibility. It's we call it
  1144. the quick fire round. So, basically
  1145. we ask you a question and you can just
  1146. tell us
  1147. quickly like what you prefer.,
  1148. because you have been an engineer
  1149. and you now also an executive
  1150. bit of management part.
  1151. So, rapid response.
  1152. What do you think the best asset for
  1153. frequency control? a 10 MW battery or 10
  1154. MW hydro turbine?
  1155. I know I don't know the capacity,
  1156. the storage, so if it's 10 MW hours or
  1157. thousands.
  1158. So,
  1159. I would take the hydro.
  1160. Very Austrian.
  1161. if you would have 1 billion euros, how
  1162. would you invest it to accelerate energy
  1163. transition or promote more flexibility?
  1164. I think I would separate it.
  1165. I think a grid still plays a huge role
  1166. to enable or make all of this happen.
  1167. And digitalization, that I think where
  1168. all of this flexibility management
  1169. system also comes to into place.
  1170. And
  1171. yeah, also some some
  1172. new missing technologies like batteries
  1173. I think
  1174. could be a game-changer, but with
  1175. digitalization also to go more
  1176. down into smaller flexibilities, maybe
  1177. even the household, e-mobility,
  1178. storage system, PV and so on households.
  1179. So, flexibility in all levels. In all
  1180. levels I think then it the billion is
  1181. gone.
  1182. In which month of the year the grid
  1183. is getting tested the most?
  1184. The summer months of August when
  1185. all the wind and the solar aid pumping
  1186. it up or in January when
  1187. it's a dark month, nothing is happening?
  1188. Yeah, I think
  1189. January is more difficult.
  1190. Yeah, because I think you always can
  1191. curtail easy, but when you have no
  1192. renewable generation then it can be
  1193. tough.
  1194. Picasso or Marie, which acronym gives
  1195. you more headaches? Headaches.
  1196. I had some headaches and I was not the
  1197. only one I think back then when we
  1198. implemented in our project.
  1199. I would say Picasso because the
  1200. Picasso go live was first. That was a
  1201. little bit harder in the beginning.
  1202. Yeah, and you already shared your
  1203. apprehension about
  1204. AI. Not just apprehension, but also like
  1205. good concern. But what do you think
  1206. about AI and energy? It's It's still
  1207. under utilized or
  1208. it's over hyped?
  1209. Yeah, I think right now it's under
  1210. utilized. We in CyberGrid and EVN I see
  1211. very active move movements into finding
  1212. new use cases and already
  1213. paint a very futuristic picture and
  1214. and like every idea is allowed and
  1215. and this is
  1216. impressive and none of us is knows in
  1217. which direction it goes. But I think
  1218. really in the daily processes we use
  1219. it a lot
  1220. too less in general and also in the
  1221. energy sector. In your role now if
  1222. you see the frequency dropping to 49.5,
  1223. what is the first thing you do? Maybe I
  1224. would call the old colleagues in the
  1225. trade room
  1226. to ask
  1227. What are the what's going on, but they
  1228. don't need my disturbance.
  1229. no, maybe I will try to see which
  1230. flexibility we have in the market and
  1231. and try to push more.
  1232. Which fire fighter, right? Yeah, that's
  1233. the
  1234. maybe the fire that activates the fire
  1235. fighter instinct, right? To press
  1236. and run direction to help the
  1237. frequency. Yeah, and also Philip
  1238. from your role in being part of
  1239. major utility here, you have like
  1240. the more clear picture of how the grid
  1241. is going to look like in let's say 7 8
  1242. years from now. How do you see in 2035
  1243. the great looking in Austria and also
  1244. all over Europe? I think a little bit
  1245. more hybrid than it is today, but I
  1246. think if you really want to make
  1247. investments into transformers and so on,
  1248. that's you need to do it today that you
  1249. have them in 10 years all what you need.
  1250. So I think we will still notice it
  1251. that that's our grid what we know now in
  1252. 35 years it would it will look similar,
  1253. but it definitely needs an investments
  1254. it need to be smarter, more hybrid, more
  1255. interconnected. Well, we're going to see
  1256. this video 10 years after and see
  1257. how it goes. Yeah,
  1258. just to finalize which advice would you
  1259. give to new electrical engineers for
  1260. example who who are entering this sector
  1261. or new professionals in energy in
  1262. general? Yeah, that's more
  1263. on the social skill level I would say,
  1264. but stay hungry, stay foolish that was
  1265. Steve Jobs back then said, but I think
  1266. to bring in the your own energy, your
  1267. own motivation, your your interest and
  1268. and your strength and I think if if if
  1269. if you're young and you're active,
  1270. motivated
  1271. you will find your way even if you can
  1272. cannot imagine where all of this then
  1273. leads to and I think that's very
  1274. important to
  1275. concentrate mainly on yourself and your
  1276. active
  1277. motivation what you can bring in and
  1278. and not
  1279. come into
  1280. the environment you are and adapt
  1281. to the environment. So bring in yourself
  1282. and your strengths and then you
  1283. will find your own way.
  1284. Amazing. on that note, thank you
  1285. very much Philip.
  1286. Finally, it was
  1287. you came, but it was
  1288. it was worth all the wait that we had
  1289. to do. Yeah, we knew it.
  1290. Yes, thank you very much. I really had
  1291. fun. Yeah, amazing. So, thank you for
  1292. coming again, Philip, and thanks to the
  1293. community for watching the episode, and
  1294. stay tuned for more episodes, and keep
  1295. tuned for the forum second and third of
  1296. June. So, see you. Ciao, ciao.